EQUATIONS can at times be tedious to understand

Size: px
Start display at page:

Download "EQUATIONS can at times be tedious to understand"

Transcription

1 DIGITAL IMAGE PROCESSING COURSE PROJECT, JUNE Augmented Reality Equation Plotte Salman Naqvi, Uzai Sikoa Digital Image Pocessing EE 368/ CS 232 Abstact Gaphical Visualization of an equation can often help us bette undestand undelying tends of eal vaiables. Moeove, visualizing it quickly as we come acoss it in a jounal aticle o textbook is often desied in ou scholaly eadings. Typing out that complex equation linealy into a numeical computing envionment, e.g. MATLAB, can be timeconsuming and hassle-pone. With the high-esolution camea, fast-pocessing, and cheap cellphones available to us, it is highly beneficial to have an equation ecognize and gaph plotte application in ou phones. Such a nifty tool can help younge students woking on building thei intuition to be able to visualize simple equations quickly in thei thoughts. Hee, we ty to develop such a tool fo the Andoid platfom. We leveage the OpenCV Andoid libay to pefom client side image pocessing and MATLAB Image Pocessing tool to pefom sleek mathematical chaacte template matching and equation fomation. The application can ead in any pinted equation and plots on sceen the gaph of an explicit equation composed of exponentials, polynomials, tigonometic functions and simple deivatives and integals. Keywods Recusive Segmentation, Binaization, Chaacte Recognition, Tesseact, Template Matching, Keypoint and Edge Detection, Andoid Development. 1 INTRODUCTION EQUATIONS can at times be tedious to undestand and moe so, difficult to quickly visualize the undelying tend of. The situation wosens if an equation is a complicated mixtue of polynomials, exponentials, and tigonometic functions, involving algebaic factions and exponents. Fotunately, moden numeical computation scipting languages, such as MATLAB, can help us quickly visualize such equations. Howeve, in case of limited access to MATLAB o to avoid witing out lage equations linealy such that they ae undestandable by the compute, we need a tool that can snap a pictue of the equation pinted on a pape and in nea eal-time pocess and etun the associated gaph on a cell phone sceen. Moeove, the gaph eading pocess can be made inteactive again by using the image pocessing techniques, and the use can view the gaph at diffeent pespective and ead values off of it. This entie pocess involves a numbe of image pocessing challenges. Fist one is the mathematical equation ecognition. Text ecognition has been widely studies in the past and efficient hand witing U. Sikoa, and S. Naqvi ae with the Depatment of Electical Engineeing, Stanfod Univesity, Stanfod, CA, usikoa, shnaqvi@stanfod.edu Manuscipt completed June 5, 2013 detection is still an aea of active eseach. On the topic of mathematical chaactes ecognition, a numbe of suvey papes published in the past decade o so, [5] and [6], sum up the developments quite well. Moe ecently, thee has been even geate wok being done in the aea of intelligent ecognition and etieval of mathematical expessions [3]. The second pocess involves a fast axes detection and plotting with pespective and skew coection. We implement these two main segments of image pocessing to develop a complete use-oiented cell phone application which can achieve its goal efficiently and vey fast. We talk about the algoithm flow chat in the next section. 2 APPLICATION FLOWCHART Ou equation plotte tool is oganized into a seies of states that ae called into activity depending on which action needs to be completed in the wokflow. An oveview of the states is shown in Figue 1. The application stats in the Captue Equation mode. In this stage, the use captues the image of the equation they would like to plot. The use can etake the image and theefoe, the state estats. Othewise, the application moves into the Equation Analysis stage. Hee, the image is sent to the seve unning MATLAB. Image segmentation, and math chaacte ecognition pat of the image

2 2 DIGITAL IMAGE PROCESSING COURSE PROJECT, JUNE 2013 pocessing takes place at this stage. The application then eceives the conveted equation sting and displays it fo the use next in the Quey Use stage. At this stage, the use has the option to edit the sting and/o appove it fo the seve to etun the coodinates vectos fo the plot. The application then moves into the Axis Detection state. Hee, the emaining of the image pocessing takes place including the canny edge detection, dilation, keypoint detection, and axis otation estimation. Afte the application has analyzed the camea peview fames and detected the coodinate axes on which to plot, we move on to the next state, Plot Points. At this point, the application calculates the location of the coodinates on the phone sceen, daws the points, and etuns to the Axis Detection state. Duing this loop between Axis Detection and Plot Points, the use has the capability to stat a new equation pocess and thus the use can etun to the Captue Equation state. The state machine appoach used in this application geatly simplifies the integation of diffeent functions and pocesses as needed. a PHP scipt to upload the image data to a file at the seve. This scipt then waits fo the MATLAB to geneate an output file at which stage, it etuns the contents of that file as an http sting. We also ensue that the image is not bigge than 640x480 pixels which limits the file size to unde 175kB fo fast cell to seve communication. We also equie use to ensue that only the equation is captued by the image and no suounding clutte (text, othe equations, etc) ae included in the image. This ensues that all egions segmented by ou algoithm ae actually pat of the equation. We can also include the functionality of allowing the use to cop the image afte taking the pictue so that only the desied equation aea is extacted and sent to the seve. Figue 2 shows an example captued image which we will use as a efeence figue thoughout this pape. Equation Analysis Captue Equation Quey Use Fig. 2: Image captued by the Andoid phone 3.2 Plot Points Axis Detection Fig. 1: The state machine diagam fo the wokflow of the Andoid application S ERVER S IDE Image Captue The pocess begins with the image captue by the cellula phone at the client s end. Once a decently focused, unifomly lit and closed up view pictue of an math equation has been taken, use posts the images to the MATLAB seve. This post initiates Binaization We need to binaize the image well so that the only egions in the image ae the math chaactes and then we can segment each individually. Otsu s method can give a easonable theshold level but with faily distubed backgound, we will see alot of backgound noise getting binaized as well. One technique is to use adaptive Otsu s method which finds the local thesholds by sliding a window acoss the entie image. Howeve, fo ou ealtime application this method is vey slow. Thus we instead implement the backgound subtaction fom the image in ode to emove any non-unifom lighting effects befoe we apply Otsu s method to find a global theshold. Backgound is detemined by dilating the gay scale image with a 24x24 disc shaped stuctuing element. Howeve, a simple dilation gives blocky esult which hampes a good binaization. Thus, we

3 NAQVI et al.: AUGMENTED REALITY EQUATION PLOTTER 3 instead use a ank filte to blu out any chaactes fom the image in a way that the esulting backgound etains its smooth blending acoss pixels. Afte tial on a numbe of lighting and image captue distubance cases, we found that a ank of 160th maximum pixel woks well. This backgound subtacted image looks as good as the binaized image 4 and it is then binaized vey efficiently using Otsu s method gay level theshold. Fig. 3: Gayscale Backgound afte 160th ode Rank Filte Afte binaization, we smooth the cones of the chaactes, by an eosion opeation followed by a dilation. This is done to impove the text ecognition esults by template matching late on. We also pefom an opening opeation to emove any small salt and peppe noise that might have been picked natually duing pictue taking pocess. to ight in the binaized image. We use the vetical pojection to find out whee the white spaces ae that sepaate the chaactes. Gadient zeo cossings of the image pojection ae used to detect the egion edges, and these egions ae extacted as subimages. Howeve, ou job is not done yet, because ou equation might have factions, in which case, we will have to sepaate the numeato and denominato fist and then to pase out individual chaacte of the numeato and denominato espectively. This will theefoe equie egion sepaation in vetical dimension on each of the hoizontally sepaated egion. We also have to do so methodically so that we keep the ight ode of numeato followed by the denominato. Afte the vetical sepaation of faction, we will attempt a hoizontal sepaation again to sepaate the numeato egions. Moeove, thee can be nested factions inside the numeato that we should sepaate. In summay, we wite a ecusive algoithm that keeps on sepaating in y fist to sepaate any lines, then in x to sepaate egions hoizontally, then in y to sepaate any factions within each egion, and then it epeats until no moe sepaation in x o y is possible. Figue 5 depicts the pocess visually. Fig. 5: Segmenting the Image Recusively Fig. 4: Binaized Image afte applying Global Otsu s theshold 3.3 Recusive Image Segmentation Afte the binaization, we can clealy visually distinguish the individual math chaactes. And if the pixels ae high enough and the image was focused well without any motion bluing, each chaacte is clealy sepaated fom the othe. We then begin sepaating the chaactes hoizontally, going fom left 3.4 Exponentiation and Faction Detection Once all the egions ae sepaated and saved as single chaacte images, we iteate ove all egions and compae the y-centoids of each egion with the one peceding it and the one following it. If the y-centoid of the cuent egion is lowe than that of the peceding one, then it maks the beginning of exponentiation, and we check the beginning of exponent label in that image s data stuctue. Moeove, if the y-centoid of the cuent egion is highe than that of the peceding egion then we

4 4 DIGITAL IMAGE PROCESSING COURSE PROJECT, JUNE 2013 check the end of exponent label in that egion s data stuctue. Afte all the egions have been labeled, we individually place the ˆ (hat) opeato followed by an open paenthesis at the stat of exponent image egion, and a close paenthesis at the end of exponent image egion. 3.5 Image Popeties and Resizing It will also be useful in the chaacte ecognition phase to know a numbe of image popeties of all the segmented chaacte images. In paticula as we shall descibe late, we ae most inteested in knowing the following popeties of each egion: 1) Aspect Ratio, which is the atio of the egion s majo axis length to the mino axis length, 2) Solidity, which is the atio of the numbe of egion pixels to the numbe of pixels in the egion s convex hull, 3) Numbe of Holes in the image, which ae obtained by subtacting fom the numbe of connected components inside an image the image s Eule numbe, 4) Centoid s y-coodinate, which gives the y location of the image s cente of mass and is helpful in detemining the sepaation of chaactes into diffeent lines, 5) Fist Hole s Centoid s y-coodinate, which will be helpful in distinguishing between vaious simila chaactes like e and 6. We stoe these popeties along with the image in a data stuctue which is stoed in the chaacte image database fo late use duing the ecognition step. But befoe stoing the image, we extact its tight bounding box and then esize it to 64x64 pixel, fo small disk space usage and easy template matching late on. The esult segmented set of subimages is shown in Figue 6. Fig. 6: Segmented egions esized with exponents and factions detected pediction model be tained to pocess these. Hand witten chaactes also need to be clealy sepaated and decently consistent in shape and size with the epetitions. SVM is a supevised machine leaning algoithm developed at the National Taiwan Univesity which analyzes data input and ecognizes patten. Fo us howeve, the task is simple because ou application is tageted towads ecognizing pinted equations appeaing in a textbook, couse notes, o jounal aticles. Thus ou fist plan of attack to accomplish easonable OCR is to use an open souce, fee OCR engine, called Tesseact. 4.1 Tesseact Output Tesseact is an feely available OCR engine that was oiginally developed at the HP Labs and is cuently being suppoted by Google [1]. It is known to pefom well in genealized cases of text ecognition whee chaactes ae limited to alphabets and digits and thee is a patten of alphabet occuence modeled by Makov chains. It employs many othe tools to ecognize text than the simple template matching and it needs to be tained fo any othe chaacte set othe than standad english lettes and numbes. In ou case, since ou goal was to be able to detect, exponentials, tigonometic functions, polynomials, simple deivatives, and geneal math symbols, ou chaacte set was lage and so alien fo Tesseact that the output was eally embaassing. Moeove, we also allowed fo diffeent fonts, which all cumulatively caused Tesseact to give vey poo esults to the expession shown ealie. The esult in which y, =,ˆ, and e wee not ecognized is shown below. 72(x6(2) + 2x + 26(x))/8(x6(3) + 5x) Fig. 7: Tesseact Output to ou example image Hence, we use an in-house template matching algoithm combined with post-pocessing of esults which geneates emakably bette esults. 4 OCR, OPTICAL CHARACTER RECOGNITION We ty two main methods fo ecognizing the segmented math chaactes images and we compae thei shotcomings. As we discussed ealie, ecognizing hand-witten equations is moe challenging and equies that a Suppot Vecto Machine (SVM) 4.2 Template Matching Database Fist, we set up a template matching database by witing down all possible chaactes that we expect to see in the types of equations we would like to solve. These amount to 35 chaactes. We also allow fo equations to be witten in 5 commonly

5 NAQVI et al.: AUGMENTED REALITY EQUATION PLOTTER 5 used fonts, including Aial, Times New Roman, Couie, Cambia Standad, and Cambia Math. The database is fed into the algoithm in the fom of an image containing of all chaactes witten in all fonts sepaated in diffeent lines. The algoithm uns a ecusive image segmentation to quickly pase out bounding boxes of all chaactes in the pope ode, computes all the helpful image popeties as discussed ealie fo all chaactes, and then esizes the images to 64x64 templates. The pocess is extemely fast and obust, and makes the template matching database easily expandable. Figue 8 shows the image used to populate the database. Fig. 8: Binaized image of the 35 chaactes with 5 font styles fed into the algoithm fo foming the template database Afte having the database setup, we use the segmented subimages fom the equation image to find the best matched template in the database, which has an associated ASCII chaacte also stoed in the database. Since both the template images and the equation segmented images ae of size 64x64 pixels, it is easie to compute thei matix dot poduct. The template giving the maximum dot poduct is taken to be the best match. We show the template matching text ecognition esult fo ou example case below. y = (xˆ(2) + 2x + 2ˆ(x))xe(xˆ(s) + 5x) Fig. 9: Template Matching Output to ou example Smat Chaacte Fixing based on Image Popeties We notice that cetain chaactes ae matched incoectly to a template in the database. In paticula, x was mistaken with a / (fowad slash), while 3 was mistaken with an s. This confusion between vaious chaactes can esult if the chaactes appeaing in the image ae skewed (skew can esult duing image captue), pixelated (pixelation can esult if fame was not zoomed in enough), blued (due to instability duing camea opeation), o pooly captued due to impope lighting conditions. Howeve, we can tackle with a decent amount of confusion by futhe filteing false positives using template and segmented images egion popeties. Fist, we an the algoithm on moe than a dozen images of diffeent equations and font styles, and unde diffeent lighting, zoom, and skew conditions. We figue out what chaacte pais ae most likely to be confused and then apply smat chaacte fixing. This is shown in the Confusion Matix plotted in Figue 10: Recognized Chaactes Database Chaactes. o c at n i s l e ( ) x y πd ^= - +* / / *+ - =^ dπ y x ( ) e l s i n t a c o. Fig. 10: Confusion Matix manually constucted by finding false positive in ove 2 dozen cases Confusion matix would keep on becoming moe dense and matue as we apply the ecognition on moe and moe divese cases. And as it populates futhe, we will get moe infomation about emedies to adapt in coecting false positives. Based on the cuent state of the confusion matix, we adapted moe than 2 dozen post-ecognition checks to coect the false esults. Fo instance, 8 is the biggest culpit which not so supisingly gets matched with many chaacte images. Thus, we deals with its set of confusion pai as follows: If the ecognized chaacte is 8, but the oiginal image has 1 hole, then if the y-centoid of the hole is above the y-centoid of the egion then ecognize it as e, othewise ecognize it

6 6 DIGITAL IMAGE PROCESSING COURSE PROJECT, JUNE 2013 as 6, and if thee wee no holes in the oiginal image and the AR is little bit highe than than of 8, then ecognize it as 5, othewise in case of small AR, ecognize it as s. With smat fixing in place of majo chaactes such as 8, (, ), 6, l, etc, ou ecognition esults impove as shown below. Now, x is coectly eplaced with /, but s was still wongly eplaced with a 5. We would need to acquie moe image featues to distinguish such simila chaactes in the futue. Fo now, we get the emaining poblems with the ecognition fixed manually by sending the ecognized equation shown below back to the use. y = (xˆ(2) + 2x + 2ˆ(x))/e(xˆ(5) + 5x) Fig. 11: Recognized Equation afte Smat Fixing 4.3 Equation pasing in MATLAB Recognizable Fomat Afte use coects the equation and sends it back, ou sting pasing algoithm kicks in to put the standad lineaized equation into a fom that is ecognized by a MATLAB o Mathematica. Majo changes to the sting in this stage includes: adding. befoe the multiplication, division, o exponentiation with a independent vaiable, adding multiplication sign befoe the paentheses, eplacing e ( with an exp(, d/dx( with dif f( fo diffeentiation, etc. Ready to solve equation looks like this: y = ((x.ˆ(2) + 2. x + 2.ˆ(x)))./(exp(x.ˆ(5) + 5. x)); Fig. 12: MATLAB Recognizable equation as Retuned by the Pasing algoithm MATLAB solves the equation fo a given ange of x values, which use can also specify, and intelligently chooses the y-limits to show the impotant plot featues within the ange of x, and sends back to the client, the sting of x, y coodinates and thei espective anges. 5 CLIENT SIDE 5.1 Coodinate Axis Detection Once the seve has pocessed the use coected equation chaacte sting, the use is pesented with the camea peview fame. At this stage in the Andoid application, the plot is ceated and ovelaid onto the image of the use povided coodinate axes. The application must continually locate the axes and plot the points as the image of the axes shifts and otates in the peview fame. We implemented the image pocessing on the phone using the OpenCV [2] Java API. The image pocessing is done between peview fames, and theefoe, fo the application to emain stable (and not fustating to the use), the image pocessing must conclude quickly to minimize lag. Theefoe, this adds a limitation to the complexity of the image pocessing opeations. The goal of the coodinate axes detection system is to locate the endpoint coodinates of the two axes. Fom these endpoints, the plot aea, oigin, otation, and scale of the plot can be detemined. We stat the detection pocess by fist extacting the gayscale intensity image fom the camea peview image. The Andoid camea povides images which ae in the YUV colospace. It follows that we can simply use the Y data fom the fame and set it as the gayscale image. We then feed this gayscale image into OpenCV s Canny edge detecto to extact the edges of the coodinate axes. The Canny edge detecto povides a good binaization of the image if thee ae no extaneous edges in the camea s field of view such as pinted text o stay maks. The pimay issue with allowing extaneous blobs in the image is that they might be candidates fo the possible endpoints of the two lines. These maks can be emoved afte Canny edge detection, using a egion labelling and egion popeties algoithm. Howeve, as we will see below, OpenCV s implementation of this algoithm slows down the coodinate detection too geatly to be utilized. The next stage of the detection pocess is to feed the Canny edge detected image into a closing opeation to fill in the edges. The Canny edge detecto [8] etuns an image with the outline of the axis. We use a dilation and eosion opeation implemented by OpenCV using a stuctuing element that is a 20 by 20 pixel squae. The esult is an image with a egion that connects the two sides of each of the edges. Finally, we use a Hais keypoint detecto to detect the cones in the image. Typically, this will esult in two keypoints being detected at each endpoint of the line. In addition to these points, keypoints ae detected along the edges of the lines due to iegulaities in the dawn line o due to

7 NAQVI et al.: AUGMENTED REALITY EQUATION PLOTTER 7 stay maks that wee left ove afte the binaization pocess. We detect the endpoints of the lines by just picking the outemost keypoints (keypoints with the maximum x, minimum x, maximum y, and minimum y). Theoetically, on a clean pape with no majo stay maks, this pocess will wok unless the use has otated the axis too fa. In this case, some of the outlying keypoints will be the same, and the application will not be able to plot the points. Picking these fou keypoints allows the application to calculate the position of the oigin. The oigin is detemined by finding the intesection of the two lines fomed by the fou keypoints. A simple atio of deteminants easily solves the system of two linea equations. The next step is to stat plotting points. ρ = θ + φ (3) a = cos ρ (4) b = sin ρ (5) Whee is the distance fom the oigin of the new point, ρ is the angle the point subtends fom the hoizontal diection on the phone, and a and b ae the coodinates of the point in the phone s coodinate system. L x and L y can be calculated by finding the distance between the two end keypoints and the oigin. Each point the MATLAB seve esponds with can be tansfomed using this method, with anothe check to detemine if the point is in the positive y quadants o the negative, as the calculation of ρ would be a subtaction instead of an addition. Fig. 13: The coodinate tansfomation that needs to be pefomed to plot points coectly on the phone sceen. Figue 13 shows the coodinate tansfomation that needs to occu afte the axes ae detemined to calculate the new coodinates of a point given its coodinates in the plot ceated by MATLAB. The oiginal coodinate system, (xy) gets tansfomed into the detected coodinate system epesented by (x y ). The coodinate system is scaled and otated by φ. Howeve, the point s location in the phone s coodinate system needs to be calculated in ode to daw the point accuately. This system is epesented by the dotted lines. If the length L x and L y ae povided along with the coodinates of the new points, the new coodinates can be detemined as follows: = ( al x L x ) 2 + ( ) bl 2 y (1) L y θ = actan bl y/l y al x/l x (2) 6 EXPERIMENTAL RESULTS 6.1 Equation Recognition Algoithm was tested on seveal types of equations and unde vaious illumination conditions fo taking equation s image. We see convincing esults in all these attempts. Unfotunately a pefect output is fa fom eal at this stage; mostly atleast one chaacte is ecognized wong o paentheses ae not closed. Fotunately, the use intevention at this stage solves the emaining issues. Fo ou detected esults, we exploe futhe the eason of getting poo template matching fo cetain types of equations. Fo ou efeence example, we plot a histogam of the distibution of matched chaactes to the diffeent font styles in Figue 14. One not so obvious, but easily justifiable obsevation, is that chaactes get wongly matched to Aial font style alot and hadly to the Times New Roman. This is to be expected because Aial font chaactes have no complicated edges and have most thick egions. Ou smat chaacte fixing jumps in to the escue and we get eally got post-pocessing esults. Figue 15 shows the esults of the algoithm applied to a numbe of diffeent equation images and the coesponding esults. We notice that most bad esults come in when we ae tying to match chaacte pais: s, 8, e, c; i, l, t, ; and 6, (, 9, ) etc. Again, as seen in the figue; most matching is to the Aial font style chaactes, although all the images used had

8 8 DIGITAL IMAGE PROCESSING COURSE PROJECT, JUNE Distibution of Chaacte Matching to Fonts Fequency ation, and Hais keypoint detecto, was chosen afte implementing seveal diffeent appoaches and examining the tadeoff between quality of detection and pocessing time. Since the eal time plotting aspect of the application needed to be pocessed with minimal lag time fom the actual peview fames, the appoach we would choose would have to be both quick and effective enough l n h d Mat bia St Roma Couie a i b Cam es New Cam Tim Aia Fig. 14: Distibution of Chaactes matched to the diffeent fonts in the Template Database equations witten oiginally in the Cambia Math style. Smat Chaacte Fixing: Distibution of Chaacte Matching to Fonts 3 Fequency y=e(.5x)c0s(4x) Template Matching: y=8(.5x)608(4x9 0 h Mat ia Std oman Couie bia Camb New R Cam es Tim l Aia Distibution of Chaacte Matching to Fonts 5 Template Matching: Fequency 4 y=2x3+x y=2x3+x2+4 0 Cam h td Mat oman Couie ia S bia Camb New R es Tim 6 Template Matching: 5 Fequency y=sin(c0s(.5x)) l Aia Distibution of Chaacte Matching to Fonts y=81n(6086.5x9) 1 0 h td Mat oman Couie ia S bia Camb New R Cam es Tim 25 l Aia Distibution of Chaacte Matching to Fonts y=81n6x^(2)+2x+2^(x)0* an(2x9^(-1x8^(6x^(3)+5x9) Fequency y=sin(x^(2)+2x+2^(x))*tan(2x)^(-1/e^((x^(3)+5x)) 0 h td Mat oman Couie ia S bia Camb New R Cam es Tim l Aia Fig. 15: Results of applying OCR on diffeent types of equations 6.2 Coodinate Axis Detection The appoach that we decided to use fo the final application using a Canny edge detecto, close ope- Fig. 16: Table of pocessing times as measued by the Andoid OS duing an axis detection and plot points cycle. Figue 16 shows the timing esults fo the thee appoaches we tested. Each test was pefomed using the same hand dawn coodinate axes to be detected. The fist appoach used backgound subtaction by subtacting the oiginal gayscale image with a dilated vesion. The flat dilation was pefomed using a stuctuing element shaped like a cicle with a diamete of 25 pixels. Then adaptive thesholding was used with a block size of 10, and the lagest egion was kept. The endpoints of the lines could be detemined by finding the outlying bight pixels. This pocessing was completed in 1320ms on the Motoola DROID phone. The second appoach we tested was using the Canny edge detecto with an OpenCV function that povided the coodinates of the endpoints of the lines it detected. This method completed in 2632ms. Finally, we tested the cuent appoach (explained in the Methods above). This pefomed in 978ms. All thee appoaches wee able to detect the axes. Howeve, the lage delays caused by the two fome appoaches pointed to using the last appoach as the final appoach. Figue 17 shows the oiginal camea fame containing the coodinate axis to be detected. Afte pefoming the Canny edge detection algoithm, the output image is shown in Figue 18. This image shows the detected edges as a line of pixels. As can be seen, the binaization pocess also emoves

9 NAQVI et al.: AUGMENTED REALITY EQUATION PLOTTER 9 Fig. 20: Hais keypoints detected Fig. 17: Oiginal gayscale camea peview. Fig. 18: Output of the Canny edges algoithm. Fig. 21: Resulting plot the backgound illumination vaiation. Figue 19 shows the esult of the closing opeation. Thee is now only one egion in the image. Next, the Hais keypoint detecto is used to find the endpoints of the axis. Figue 20 shows the keypoints that wee detected. Since the lines ae not pefectly staight, the Canny edge, and closing opeations leave small cones that the keypoint detecto picks up. This is not an issue because of ou outlie keypoint selecto. Figue 21 is a photogaph of the phone sceen displaying the plot that was geneated using the points etuned by the seve. As shown, the plot displays coectly ovelaid on the dawn coodinate axes Fig. 19: Afte applying close opeation. D ISCUSSION Coodinate Axis Detection The esults fo the axis detection appoach ae pomising. Fo elatively clean images, the pocess woks as expected. Howeve, the appoach does not take into account outlying egions and theefoe the axes would not be accuately detected. The appoach of egion filteing with the Canny edge detecto and Hais keypoint detecto could emedy this issue. Howeve, the delay additional pocessing intoduces is an issue that must be dealt with. Fo elatively newe phones, the pocessing powe upgade possibly could limit the time lag caused by the pocessing and moe sophisticated detection could be pefomed.

10 10 DIGITAL IMAGE PROCESSING COURSE PROJECT, JUNE 2013 If futue inteactivity featues of the application ae taken into consideation, the appoach used hee fo axis detection might be beneficial. It is both fast and accuate enough to detect the axis, but in addition, keypoints detected within the bounding box aound the plot aea defined by the limits of the axis ae not possible axis endpoints. Theefoe, maks made in the inteio of the plot do not affect the detection. An additional featue of the application could take advantage of this by allowing the use to daw points in the inteio of the plot aea. These could be detected as egions o possible keypoints that do not fall nea the axis. The coodinates of these use defined points could be povided fo futhe plot infomation. 8 FUTURE WORK This tool has many useful extensions and many beneficial applications. As a fist next step, we can wok on making the eal-time plotting of the gaph insensitive to pespective distotion along with skewness. We would also like to make the OCR insensitive to image captue, in paticula, making it insensitive to skew and pespective distotion. We can use the RANSAC MATLAB Toolbox fom the Gaphics and Media Lab of Lomonosov Moscow State Univesity [7] to compute RANdom SAmple Consensus (RANSAC) that ts to the top and the baseline of the equation in the image. In this way we can find the coected bounding box of the image and apply the pojective affine tansfom to deskew and pespective compensate the image. Also, to impove the chaacte ecognition in the cuent setup, we would like to make a moe detailed confusion matix and encode moe classifies fo minimizing false positives. We would also like to allow fo including the handwitten equation ecognition capability by taining an SVM as discussed ealie. Anothe goal of the poject that we stopped shot of achieving was to add the inteactive plot eading capability, whee use can put a dot within the axes pemises on the pape, and the tool can ovelay on the sceen the eading of the gaph at that location. Finally, being able to visualize 3D gaphs would be a gem of an accomplishment fo this type of a poject. seve. We get eally nice esults with elatively simple template matching scheme with elatively low numbe of post-matching image classifies. ACKNOWLEDGMENTS The authos would like to thank Pof. Bend Giod fo teaching them nifty image pocessing tools to solve many useful eal-wold poblems. They ae also indebted to the helpful Teaching Assistants, David Chen and Matt Yu, who was eve-eady to discuss poject ideas and povide citical feedback. Moeove, vaious intellectual discussions held with Sam Tsai, gaduate student at the SCIEN Lab, wee eally mind-opening and helpful fo cafting pathway fo chaacte ecognition. They also extend thei thanks to fiend, Patick Blaes, fo helping out in setting up the seve. REFERENCES [1] Tesseact Optical Chaacte Recognition Engine. Available: [2] Itseez, Andoid OpenCV, [Online]. Available: [3] R. Zanibbi and D. Blostein, Recognition and etieval of mathematical expessions, Intenational Jounal on Document Analysis and Recognition (IJDAR), vol. 15, pp , [4] S. S. Tsai, H. Chen, D. Chen, R. Vedantham, R. Gzeszczuk and B. Giod, Mobile visual seach using image and text featues, in Signals, Systems and Computes (ASILO- MAR), 2011 Confeence Recod of the Foty Fifth Asiloma Confeence on, [5] K. Chan and D. Yeung, Mathematical Expession Recognition: A Suvey, Intl J. Document Analysis and Recognition, vol. 3, no. 1, pp. 3-15, Aug [6] D. Blostein and A. Gbavec, Recognition of Mathematical Notation, Handbook of Chaacte Recognition and Document Image Analysis, pp , Wold Scientific, [7] Gaphics and Media Lab RANSAC Matlab Toolbox. Available: machineleaning/ansactoolbox [8] Canny, J., A Computational Appoach To Edge Detection, IEEE Tans. Patten Analysis and Machine Intelligence, 8(6):679698, CONCLUSION We have successfully showed that nea eal-time pinted explicit mathematical equation plotting can be achieved using Andoid OpenCV and MATLAB

Journal of World s Electrical Engineering and Technology J. World. Elect. Eng. Tech. 1(1): 12-16, 2012

Journal of World s Electrical Engineering and Technology J. World. Elect. Eng. Tech. 1(1): 12-16, 2012 2011, Scienceline Publication www.science-line.com Jounal of Wold s Electical Engineeing and Technology J. Wold. Elect. Eng. Tech. 1(1): 12-16, 2012 JWEET An Efficient Algoithm fo Lip Segmentation in Colo

More information

Detection and Recognition of Alert Traffic Signs

Detection and Recognition of Alert Traffic Signs Detection and Recognition of Alet Taffic Signs Chia-Hsiung Chen, Macus Chen, and Tianshi Gao 1 Stanfod Univesity Stanfod, CA 9305 {echchen, macuscc, tianshig}@stanfod.edu Abstact Taffic signs povide dives

More information

Image Enhancement in the Spatial Domain. Spatial Domain

Image Enhancement in the Spatial Domain. Spatial Domain 8-- Spatial Domain Image Enhancement in the Spatial Domain What is spatial domain The space whee all pixels fom an image In spatial domain we can epesent an image by f( whee x and y ae coodinates along

More information

Color Correction Using 3D Multiview Geometry

Color Correction Using 3D Multiview Geometry Colo Coection Using 3D Multiview Geomety Dong-Won Shin and Yo-Sung Ho Gwangju Institute of Science and Technology (GIST) 13 Cheomdan-gwagio, Buk-ku, Gwangju 500-71, Republic of Koea ABSTRACT Recently,

More information

Prof. Feng Liu. Fall /17/2016

Prof. Feng Liu. Fall /17/2016 Pof. Feng Liu Fall 26 http://www.cs.pdx.edu/~fliu/couses/cs447/ /7/26 Last time Compositing NPR 3D Gaphics Toolkits Tansfomations 2 Today 3D Tansfomations The Viewing Pipeline Mid-tem: in class, Nov. 2

More information

Segmentation of Casting Defects in X-Ray Images Based on Fractal Dimension

Segmentation of Casting Defects in X-Ray Images Based on Fractal Dimension 17th Wold Confeence on Nondestuctive Testing, 25-28 Oct 2008, Shanghai, China Segmentation of Casting Defects in X-Ray Images Based on Factal Dimension Jue WANG 1, Xiaoqin HOU 2, Yufang CAI 3 ICT Reseach

More information

An Unsupervised Segmentation Framework For Texture Image Queries

An Unsupervised Segmentation Framework For Texture Image Queries An Unsupevised Segmentation Famewok Fo Textue Image Queies Shu-Ching Chen Distibuted Multimedia Infomation System Laboatoy School of Compute Science Floida Intenational Univesity Miami, FL 33199, USA chens@cs.fiu.edu

More information

4.2. Co-terminal and Related Angles. Investigate

4.2. Co-terminal and Related Angles. Investigate .2 Co-teminal and Related Angles Tigonometic atios can be used to model quantities such as

More information

Optical Flow for Large Motion Using Gradient Technique

Optical Flow for Large Motion Using Gradient Technique SERBIAN JOURNAL OF ELECTRICAL ENGINEERING Vol. 3, No. 1, June 2006, 103-113 Optical Flow fo Lage Motion Using Gadient Technique Md. Moshaof Hossain Sake 1, Kamal Bechkoum 2, K.K. Islam 1 Abstact: In this

More information

Lecture # 04. Image Enhancement in Spatial Domain

Lecture # 04. Image Enhancement in Spatial Domain Digital Image Pocessing CP-7008 Lectue # 04 Image Enhancement in Spatial Domain Fall 2011 2 domains Spatial Domain : (image plane) Techniques ae based on diect manipulation of pixels in an image Fequency

More information

A Two-stage and Parameter-free Binarization Method for Degraded Document Images

A Two-stage and Parameter-free Binarization Method for Degraded Document Images A Two-stage and Paamete-fee Binaization Method fo Degaded Document Images Yung-Hsiang Chiu 1, Kuo-Liang Chung 1, Yong-Huai Huang 2, Wei-Ning Yang 3, Chi-Huang Liao 4 1 Depatment of Compute Science and

More information

Illumination methods for optical wear detection

Illumination methods for optical wear detection Illumination methods fo optical wea detection 1 J. Zhang, 2 P.P.L.Regtien 1 VIMEC Applied Vision Technology, Coy 43, 5653 LC Eindhoven, The Nethelands Email: jianbo.zhang@gmail.com 2 Faculty Electical

More information

Positioning of a robot based on binocular vision for hand / foot fusion Long Han

Positioning of a robot based on binocular vision for hand / foot fusion Long Han 2nd Intenational Confeence on Advances in Mechanical Engineeing and Industial Infomatics (AMEII 26) Positioning of a obot based on binocula vision fo hand / foot fusion Long Han Compute Science and Technology,

More information

Topic -3 Image Enhancement

Topic -3 Image Enhancement Topic -3 Image Enhancement (Pat 1) DIP: Details Digital Image Pocessing Digital Image Chaacteistics Spatial Spectal Gay-level Histogam DFT DCT Pe-Pocessing Enhancement Restoation Point Pocessing Masking

More information

A Memory Efficient Array Architecture for Real-Time Motion Estimation

A Memory Efficient Array Architecture for Real-Time Motion Estimation A Memoy Efficient Aay Achitectue fo Real-Time Motion Estimation Vasily G. Moshnyaga and Keikichi Tamau Depatment of Electonics & Communication, Kyoto Univesity Sakyo-ku, Yoshida-Honmachi, Kyoto 66-1, JAPAN

More information

Spiral Recognition Methodology and Its Application for Recognition of Chinese Bank Checks

Spiral Recognition Methodology and Its Application for Recognition of Chinese Bank Checks Spial Recognition Methodology and Its Application fo Recognition of Chinese Bank Checks Hanshen Tang 1, Emmanuel Augustin 2, Ching Y. Suen 1, Olivie Baet 2, Mohamed Cheiet 3 1 Cente fo Patten Recognition

More information

Towards Adaptive Information Merging Using Selected XML Fragments

Towards Adaptive Information Merging Using Selected XML Fragments Towads Adaptive Infomation Meging Using Selected XML Fagments Ho-Lam Lau and Wilfed Ng Depatment of Compute Science and Engineeing, The Hong Kong Univesity of Science and Technology, Hong Kong {lauhl,

More information

(a, b) x y r. For this problem, is a point in the - coordinate plane and is a positive number.

(a, b) x y r. For this problem, is a point in the - coordinate plane and is a positive number. Illustative G-C Simila cicles Alignments to Content Standads: G-C.A. Task (a, b) x y Fo this poblem, is a point in the - coodinate plane and is a positive numbe. a. Using a tanslation and a dilation, show

More information

MapReduce Optimizations and Algorithms 2015 Professor Sasu Tarkoma

MapReduce Optimizations and Algorithms 2015 Professor Sasu Tarkoma apreduce Optimizations and Algoithms 2015 Pofesso Sasu Takoma www.cs.helsinki.fi Optimizations Reduce tasks cannot stat befoe the whole map phase is complete Thus single slow machine can slow down the

More information

A New and Efficient 2D Collision Detection Method Based on Contact Theory Xiaolong CHENG, Jun XIAO a, Ying WANG, Qinghai MIAO, Jian XUE

A New and Efficient 2D Collision Detection Method Based on Contact Theory Xiaolong CHENG, Jun XIAO a, Ying WANG, Qinghai MIAO, Jian XUE 5th Intenational Confeence on Advanced Mateials and Compute Science (ICAMCS 2016) A New and Efficient 2D Collision Detection Method Based on Contact Theoy Xiaolong CHENG, Jun XIAO a, Ying WANG, Qinghai

More information

Lecture 27: Voronoi Diagrams

Lecture 27: Voronoi Diagrams We say that two points u, v Y ae in the same connected component of Y if thee is a path in R N fom u to v such that all the points along the path ae in the set Y. (Thee ae two connected components in the

More information

A Novel Automatic White Balance Method For Digital Still Cameras

A Novel Automatic White Balance Method For Digital Still Cameras A Novel Automatic White Balance Method Fo Digital Still Cameas Ching-Chih Weng 1, Home Chen 1,2, and Chiou-Shann Fuh 3 Depatment of Electical Engineeing, 2 3 Gaduate Institute of Communication Engineeing

More information

IP Network Design by Modified Branch Exchange Method

IP Network Design by Modified Branch Exchange Method Received: June 7, 207 98 IP Netwok Design by Modified Banch Method Kaiat Jaoenat Natchamol Sichumoenattana 2* Faculty of Engineeing at Kamphaeng Saen, Kasetsat Univesity, Thailand 2 Faculty of Management

More information

Module 6 STILL IMAGE COMPRESSION STANDARDS

Module 6 STILL IMAGE COMPRESSION STANDARDS Module 6 STILL IMAE COMPRESSION STANDARDS Lesson 17 JPE-2000 Achitectue and Featues Instuctional Objectives At the end of this lesson, the students should be able to: 1. State the shotcomings of JPE standad.

More information

2. PROPELLER GEOMETRY

2. PROPELLER GEOMETRY a) Fames of Refeence 2. PROPELLER GEOMETRY 10 th Intenational Towing Tank Committee (ITTC) initiated the pepaation of a dictionay and nomenclatue of ship hydodynamic tems and this wok was completed in

More information

Fifth Wheel Modelling and Testing

Fifth Wheel Modelling and Testing Fifth heel Modelling and Testing en Masoy Mechanical Engineeing Depatment Floida Atlantic Univesity Boca aton, FL 4 Lois Malaptias IFMA Institut Fancais De Mechanique Advancee ampus De lemont Feand Les

More information

Obstacle Avoidance of Autonomous Mobile Robot using Stereo Vision Sensor

Obstacle Avoidance of Autonomous Mobile Robot using Stereo Vision Sensor Obstacle Avoidance of Autonomous Mobile Robot using Steeo Vision Senso Masako Kumano Akihisa Ohya Shin ichi Yuta Intelligent Robot Laboatoy Univesity of Tsukuba, Ibaaki, 35-8573 Japan E-mail: {masako,

More information

Goal. Rendering Complex Scenes on Mobile Terminals or on the web. Rendering on Mobile Terminals. Rendering on Mobile Terminals. Walking through images

Goal. Rendering Complex Scenes on Mobile Terminals or on the web. Rendering on Mobile Terminals. Rendering on Mobile Terminals. Walking through images Goal Walking though s -------------------------------------------- Kadi Bouatouch IRISA Univesité de Rennes I, Fance Rendeing Comple Scenes on Mobile Teminals o on the web Rendeing on Mobile Teminals Rendeing

More information

Assessment of Track Sequence Optimization based on Recorded Field Operations

Assessment of Track Sequence Optimization based on Recorded Field Operations Assessment of Tack Sequence Optimization based on Recoded Field Opeations Matin A. F. Jensen 1,2,*, Claus G. Søensen 1, Dionysis Bochtis 1 1 Aahus Univesity, Faculty of Science and Technology, Depatment

More information

TESSELLATIONS. This is a sample (draft) chapter from: MATHEMATICAL OUTPOURINGS. Newsletters and Musings from the St. Mark s Institute of Mathematics

TESSELLATIONS. This is a sample (draft) chapter from: MATHEMATICAL OUTPOURINGS. Newsletters and Musings from the St. Mark s Institute of Mathematics TESSELLATIONS This is a sample (daft) chapte fom: MATHEMATICAL OUTPOURINGS Newslettes and Musings fom the St. Mak s Institute of Mathematics James Tanton www.jamestanton.com This mateial was and can still

More information

Title. Author(s)NOMURA, K.; MOROOKA, S. Issue Date Doc URL. Type. Note. File Information

Title. Author(s)NOMURA, K.; MOROOKA, S. Issue Date Doc URL. Type. Note. File Information Title CALCULATION FORMULA FOR A MAXIMUM BENDING MOMENT AND THE TRIANGULAR SLAB WITH CONSIDERING EFFECT OF SUPPO UNIFORM LOAD Autho(s)NOMURA, K.; MOROOKA, S. Issue Date 2013-09-11 Doc URL http://hdl.handle.net/2115/54220

More information

Controlled Information Maximization for SOM Knowledge Induced Learning

Controlled Information Maximization for SOM Knowledge Induced Learning 3 Int'l Conf. Atificial Intelligence ICAI'5 Contolled Infomation Maximization fo SOM Knowledge Induced Leaning Ryotao Kamimua IT Education Cente and Gaduate School of Science and Technology, Tokai Univeisity

More information

MULTI-TEMPORAL AND MULTI-SENSOR IMAGE MATCHING BASED ON LOCAL FREQUENCY INFORMATION

MULTI-TEMPORAL AND MULTI-SENSOR IMAGE MATCHING BASED ON LOCAL FREQUENCY INFORMATION Intenational Achives of the Photogammety Remote Sensing and Spatial Infomation Sciences Volume XXXIX-B3 2012 XXII ISPRS Congess 25 August 01 Septembe 2012 Melboune Austalia MULTI-TEMPORAL AND MULTI-SENSOR

More information

Augmented Reality. Integrating Computer Graphics with Computer Vision Mihran Tuceryan. August 16, 1998 ICPR 98 1

Augmented Reality. Integrating Computer Graphics with Computer Vision Mihran Tuceryan. August 16, 1998 ICPR 98 1 Augmented Reality Integating Compute Gaphics with Compute Vision Mihan Tuceyan August 6, 998 ICPR 98 Definition XCombines eal and vitual wolds and objects XIt is inteactive and eal-time XThe inteaction

More information

A modal estimation based multitype sensor placement method

A modal estimation based multitype sensor placement method A modal estimation based multitype senso placement method *Xue-Yang Pei 1), Ting-Hua Yi 2) and Hong-Nan Li 3) 1),)2),3) School of Civil Engineeing, Dalian Univesity of Technology, Dalian 116023, China;

More information

17/5/2009. Introduction

17/5/2009. Introduction 7/5/9 Steeo Imaging Intoduction Eample of Human Vision Peception of Depth fom Left and ight eye images Diffeence in elative position of object in left and ight eyes. Depth infomation in the views?? 7/5/9

More information

dc - Linux Command Dc may be invoked with the following command-line options: -V --version Print out the version of dc

dc - Linux Command Dc may be invoked with the following command-line options: -V --version Print out the version of dc - CentOS 5.2 - Linux Uses Guide - Linux Command SYNOPSIS [-V] [--vesion] [-h] [--help] [-e sciptexpession] [--expession=sciptexpession] [-f sciptfile] [--file=sciptfile] [file...] DESCRIPTION is a evese-polish

More information

An Extension to the Local Binary Patterns for Image Retrieval

An Extension to the Local Binary Patterns for Image Retrieval , pp.81-85 http://x.oi.og/10.14257/astl.2014.45.16 An Extension to the Local Binay Pattens fo Image Retieval Zhize Wu, Yu Xia, Shouhong Wan School of Compute Science an Technology, Univesity of Science

More information

Frequency Domain Approach for Face Recognition Using Optical Vanderlugt Filters

Frequency Domain Approach for Face Recognition Using Optical Vanderlugt Filters Optics and Photonics Jounal, 016, 6, 94-100 Published Online August 016 in SciRes. http://www.scip.og/jounal/opj http://dx.doi.og/10.436/opj.016.68b016 Fequency Domain Appoach fo Face Recognition Using

More information

Shortest Paths for a Two-Robot Rendez-Vous

Shortest Paths for a Two-Robot Rendez-Vous Shotest Paths fo a Two-Robot Rendez-Vous Eik L Wyntes Joseph S B Mitchell y Abstact In this pape, we conside an optimal motion planning poblem fo a pai of point obots in a plana envionment with polygonal

More information

Information Retrieval. CS630 Representing and Accessing Digital Information. IR Basics. User Task. Basic IR Processes

Information Retrieval. CS630 Representing and Accessing Digital Information. IR Basics. User Task. Basic IR Processes CS630 Repesenting and Accessing Digital Infomation Infomation Retieval: Basics Thosten Joachims Conell Univesity Infomation Retieval Basics Retieval Models Indexing and Pepocessing Data Stuctues ~ 4 lectues

More information

POMDP: Introduction to Partially Observable Markov Decision Processes Hossein Kamalzadeh, Michael Hahsler

POMDP: Introduction to Partially Observable Markov Decision Processes Hossein Kamalzadeh, Michael Hahsler POMDP: Intoduction to Patially Obsevable Makov Decision Pocesses Hossein Kamalzadeh, Michael Hahsle 2019-01-02 The R package pomdp povides an inteface to pomdp-solve, a solve (witten in C) fo Patially

More information

An Optimised Density Based Clustering Algorithm

An Optimised Density Based Clustering Algorithm Intenational Jounal of Compute Applications (0975 8887) Volume 6 No.9, Septembe 010 An Optimised Density Based Clusteing Algoithm J. Hencil Pete Depatment of Compute Science St. Xavie s College, Palayamkottai,

More information

ADDING REALISM TO SOURCE CHARACTERIZATION USING A GENETIC ALGORITHM

ADDING REALISM TO SOURCE CHARACTERIZATION USING A GENETIC ALGORITHM ADDING REALISM TO SOURCE CHARACTERIZATION USING A GENETIC ALGORITHM Luna M. Rodiguez*, Sue Ellen Haupt, and Geoge S. Young Depatment of Meteoology and Applied Reseach Laboatoy The Pennsylvania State Univesity,

More information

A Mathematical Implementation of a Global Human Walking Model with Real-Time Kinematic Personification by Boulic, Thalmann and Thalmann.

A Mathematical Implementation of a Global Human Walking Model with Real-Time Kinematic Personification by Boulic, Thalmann and Thalmann. A Mathematical Implementation of a Global Human Walking Model with Real-Time Kinematic Pesonification by Boulic, Thalmann and Thalmann. Mashall Badley National Cente fo Physical Acoustics Univesity of

More information

Cardiac C-Arm CT. SNR Enhancement by Combining Multiple Retrospectively Motion Corrected FDK-Like Reconstructions

Cardiac C-Arm CT. SNR Enhancement by Combining Multiple Retrospectively Motion Corrected FDK-Like Reconstructions Cadiac C-Am CT SNR Enhancement by Combining Multiple Retospectively Motion Coected FDK-Like Reconstuctions M. Pümme 1, L. Wigstöm 2,3, R. Fahig 2, G. Lauitsch 4, J. Honegge 1 1 Institute of Patten Recognition,

More information

Any modern computer system will incorporate (at least) two levels of storage:

Any modern computer system will incorporate (at least) two levels of storage: 1 Any moden compute system will incopoate (at least) two levels of stoage: pimay stoage: andom access memoy (RAM) typical capacity 32MB to 1GB cost pe MB $3. typical access time 5ns to 6ns bust tansfe

More information

A Novel Image-Based Rendering System With A Longitudinally Aligned Camera Array

A Novel Image-Based Rendering System With A Longitudinally Aligned Camera Array EUOGAPHICS 2 / A. de Sousa, J.C. Toes Shot Pesentations A Novel Image-Based endeing System With A Longitudinally Aligned Camea Aay Jiang Li, Kun Zhou, Yong Wang and Heung-Yeung Shum Micosoft eseach, China

More information

OPTIMAL KINEMATIC SYNTHESIS OF CRANK & SLOTTED LEVER QUICK RETURN MECHANISM FOR SPECIFIC STROKE & TIME RATIO

OPTIMAL KINEMATIC SYNTHESIS OF CRANK & SLOTTED LEVER QUICK RETURN MECHANISM FOR SPECIFIC STROKE & TIME RATIO OPTIMAL KINEMATIC SYNTHESIS OF CRANK & SLOTTED LEVER QUICK RETURN MECHANISM FOR SPECIFIC STROKE & TIME RATIO Zeeshan A. Shaikh 1 and T.Y. Badguja 2 1,2 Depatment of Mechanical Engineeing, Late G. N. Sapkal

More information

Image Registration among UAV Image Sequence and Google Satellite Image Under Quality Mismatch

Image Registration among UAV Image Sequence and Google Satellite Image Under Quality Mismatch 0 th Intenational Confeence on ITS Telecommunications Image Registation among UAV Image Sequence and Google Satellite Image Unde Quality Mismatch Shih-Ming Huang and Ching-Chun Huang Depatment of Electical

More information

Survey of Various Image Enhancement Techniques in Spatial Domain Using MATLAB

Survey of Various Image Enhancement Techniques in Spatial Domain Using MATLAB Suvey of Vaious Image Enhancement Techniques in Spatial Domain Using MATLAB Shailenda Singh Negi M.Tech Schola G. B. Pant Engineeing College, Paui Gahwal Uttaahand, India- 46194 ABSTRACT Image Enhancement

More information

AUTOMATED LOCATION OF ICE REGIONS IN RADARSAT SAR IMAGERY

AUTOMATED LOCATION OF ICE REGIONS IN RADARSAT SAR IMAGERY AUTOMATED LOCATION OF ICE REGIONS IN RADARSAT SAR IMAGERY Chistophe Waceman (1), William G. Pichel (2), Pablo Clement-Colón (2) (1) Geneal Dynamics Advanced Infomation Systems, P.O. Box 134008 Ann Abo

More information

Transmission Lines Modeling Based on Vector Fitting Algorithm and RLC Active/Passive Filter Design

Transmission Lines Modeling Based on Vector Fitting Algorithm and RLC Active/Passive Filter Design Tansmission Lines Modeling Based on Vecto Fitting Algoithm and RLC Active/Passive Filte Design Ahmed Qasim Tuki a,*, Nashien Fazilah Mailah b, Mohammad Lutfi Othman c, Ahmad H. Saby d Cente fo Advanced

More information

RANDOM IRREGULAR BLOCK-HIERARCHICAL NETWORKS: ALGORITHMS FOR COMPUTATION OF MAIN PROPERTIES

RANDOM IRREGULAR BLOCK-HIERARCHICAL NETWORKS: ALGORITHMS FOR COMPUTATION OF MAIN PROPERTIES RANDOM IRREGULAR BLOCK-HIERARCHICAL NETWORKS: ALGORITHMS FOR COMPUTATION OF MAIN PROPERTIES Svetlana Avetisyan Mikayel Samvelyan* Matun Kaapetyan Yeevan State Univesity Abstact In this pape, the class

More information

A Neural Network Model for Storing and Retrieving 2D Images of Rotated 3D Object Using Principal Components

A Neural Network Model for Storing and Retrieving 2D Images of Rotated 3D Object Using Principal Components A Neual Netwok Model fo Stong and Reteving 2D Images of Rotated 3D Object Using Pncipal Components Tsukasa AMANO, Shuichi KUROGI, Ayako EGUCHI, Takeshi NISHIDA, Yasuhio FUCHIKAWA Depatment of Contol Engineeng,

More information

Slotted Random Access Protocol with Dynamic Transmission Probability Control in CDMA System

Slotted Random Access Protocol with Dynamic Transmission Probability Control in CDMA System Slotted Random Access Potocol with Dynamic Tansmission Pobability Contol in CDMA System Intaek Lim 1 1 Depatment of Embedded Softwae, Busan Univesity of Foeign Studies, itlim@bufs.ac.k Abstact In packet

More information

EYE DIRECTION BY STEREO IMAGE PROCESSING USING CORNEAL REFLECTION ON AN IRIS

EYE DIRECTION BY STEREO IMAGE PROCESSING USING CORNEAL REFLECTION ON AN IRIS EYE DIRECTION BY STEREO IMAGE PROCESSING USING CORNEAL REFLECTION ON AN IRIS Kumiko Tsuji Fukuoka Medical technology Teikyo Univesity 4-3-14 Shin-Katsutachi-Machi Ohmuta Fukuoka 836 Japan email: c746g@wisdomcckyushu-uacjp

More information

3D Hand Trajectory Segmentation by Curvatures and Hand Orientation for Classification through a Probabilistic Approach

3D Hand Trajectory Segmentation by Curvatures and Hand Orientation for Classification through a Probabilistic Approach 3D Hand Tajectoy Segmentation by Cuvatues and Hand Oientation fo Classification though a Pobabilistic Appoach Diego R. Faia and Joge Dias Abstact In this wok we pesent the segmentation and classification

More information

a Not yet implemented in current version SPARK: Research Kit Pointer Analysis Parameters Soot Pointer analysis. Objectives

a Not yet implemented in current version SPARK: Research Kit Pointer Analysis Parameters Soot Pointer analysis. Objectives SPARK: Soot Reseach Kit Ondřej Lhoták Objectives Spak is a modula toolkit fo flow-insensitive may points-to analyses fo Java, which enables expeimentation with: vaious paametes of pointe analyses which

More information

Research Article. Regularization Rotational motion image Blur Restoration

Research Article. Regularization Rotational motion image Blur Restoration Available online www.jocp.com Jounal of Chemical and Phamaceutical Reseach, 6, 8(6):47-476 Reseach Aticle ISSN : 975-7384 CODEN(USA) : JCPRC5 Regulaization Rotational motion image Blu Restoation Zhen Chen

More information

FACE VECTORS OF FLAG COMPLEXES

FACE VECTORS OF FLAG COMPLEXES FACE VECTORS OF FLAG COMPLEXES ANDY FROHMADER Abstact. A conjectue of Kalai and Eckhoff that the face vecto of an abitay flag complex is also the face vecto of some paticula balanced complex is veified.

More information

t [ Background removed

t [ Background removed HANDS-ON > HOE1 Taining 1 Pictues You decide to ceate a memoies slide show fo you siste and he husband, who wee ecently maied. You include thei high school Sweetheat Ball image, engagement and wedding

More information

Point-Biserial Correlation Analysis of Fuzzy Attributes

Point-Biserial Correlation Analysis of Fuzzy Attributes Appl Math Inf Sci 6 No S pp 439S-444S (0 Applied Mathematics & Infomation Sciences An Intenational Jounal @ 0 NSP Natual Sciences Publishing o Point-iseial oelation Analysis of Fuzzy Attibutes Hao-En hueh

More information

Input Layer f = 2 f = 0 f = f = 3 1,16 1,1 1,2 1,3 2, ,2 3,3 3,16. f = 1. f = Output Layer

Input Layer f = 2 f = 0 f = f = 3 1,16 1,1 1,2 1,3 2, ,2 3,3 3,16. f = 1. f = Output Layer Using the Gow-And-Pune Netwok to Solve Poblems of Lage Dimensionality B.J. Biedis and T.D. Gedeon School of Compute Science & Engineeing The Univesity of New South Wales Sydney NSW 2052 AUSTRALIA bbiedis@cse.unsw.edu.au

More information

Comparisons of Transient Analytical Methods for Determining Hydraulic Conductivity Using Disc Permeameters

Comparisons of Transient Analytical Methods for Determining Hydraulic Conductivity Using Disc Permeameters Compaisons of Tansient Analytical Methods fo Detemining Hydaulic Conductivity Using Disc Pemeametes 1,,3 Cook, F.J. 1 CSRO Land and Wate, ndoooopilly, Queensland The Univesity of Queensland, St Lucia,

More information

Pipes, connections, channels and multiplexors

Pipes, connections, channels and multiplexors Pipes, connections, channels and multiplexos Fancisco J. Ballesteos ABSTRACT Channels in the style of CSP ae a poeful abstaction. The ae close to pipes and connections used to inteconnect system and netok

More information

A VECTOR PERTURBATION APPROACH TO THE GENERALIZED AIRCRAFT SPARE PARTS GROUPING PROBLEM

A VECTOR PERTURBATION APPROACH TO THE GENERALIZED AIRCRAFT SPARE PARTS GROUPING PROBLEM Accepted fo publication Intenational Jounal of Flexible Automation and Integated Manufactuing. A VECTOR PERTURBATION APPROACH TO THE GENERALIZED AIRCRAFT SPARE PARTS GROUPING PROBLEM Nagiza F. Samatova,

More information

Ego-Motion Estimation on Range Images using High-Order Polynomial Expansion

Ego-Motion Estimation on Range Images using High-Order Polynomial Expansion Ego-Motion Estimation on Range Images using High-Ode Polynomial Expansion Bian Okon and Josh Haguess Space and Naval Wafae Systems Cente Pacific San Diego, CA, USA {bian.okon,joshua.haguess}@navy.mil Abstact

More information

Color Interpolation for Single CCD Color Camera

Color Interpolation for Single CCD Color Camera Colo Intepolation fo Single CCD Colo Camea Yi-Ming Wu, Chiou-Shann Fuh, and Jui-Pin Hsu Depatment of Compute Science and Infomation Engineeing, National Taian Univesit, Taipei, Taian Email: 88036@csie.ntu.edu.t;

More information

5 4 THE BERNOULLI EQUATION

5 4 THE BERNOULLI EQUATION 185 CHATER 5 the suounding ai). The fictional wok tem w fiction is often expessed as e loss to epesent the loss (convesion) of mechanical into themal. Fo the idealied case of fictionless motion, the last

More information

Extract Object Boundaries in Noisy Images using Level Set. Final Report

Extract Object Boundaries in Noisy Images using Level Set. Final Report Extact Object Boundaies in Noisy Images using Level Set by: Quming Zhou Final Repot Submitted to Pofesso Bian Evans EE381K Multidimensional Digital Signal Pocessing May 10, 003 Abstact Finding object contous

More information

UCLA Papers. Title. Permalink. Authors. Publication Date. Localized Edge Detection in Sensor Fields. https://escholarship.org/uc/item/3fj6g58j

UCLA Papers. Title. Permalink. Authors. Publication Date. Localized Edge Detection in Sensor Fields. https://escholarship.org/uc/item/3fj6g58j UCLA Papes Title Localized Edge Detection in Senso Fields Pemalink https://escholashipog/uc/item/3fj6g58j Authos K Chintalapudi Govindan Publication Date 3-- Pee eviewed escholashipog Poweed by the Califonia

More information

A Recommender System for Online Personalization in the WUM Applications

A Recommender System for Online Personalization in the WUM Applications A Recommende System fo Online Pesonalization in the WUM Applications Mehdad Jalali 1, Nowati Mustapha 2, Ali Mamat 2, Md. Nasi B Sulaiman 2 Abstact foeseeing of use futue movements and intentions based

More information

Also available at ISSN (printed edn.), ISSN (electronic edn.) ARS MATHEMATICA CONTEMPORANEA 3 (2010)

Also available at  ISSN (printed edn.), ISSN (electronic edn.) ARS MATHEMATICA CONTEMPORANEA 3 (2010) Also available at http://amc.imfm.si ISSN 1855-3966 (pinted edn.), ISSN 1855-3974 (electonic edn.) ARS MATHEMATICA CONTEMPORANEA 3 (2010) 109 120 Fulleene patches I Jack E. Gave Syacuse Univesity, Depatment

More information

A ROI Focusing Mechanism for Digital Cameras

A ROI Focusing Mechanism for Digital Cameras A ROI Focusing Mechanism fo Digital Cameas Chu-Hui Lee, Meng-Feng Lin, Chun-Ming Huang, and Chun-Wei Hsu Abstact With the development and application of digital technologies, the digital camea is moe popula

More information

Massachusetts Institute of Technology Department of Mechanical Engineering

Massachusetts Institute of Technology Department of Mechanical Engineering cm cm Poblem Massachusetts Institute of echnolog Depatment of Mechanical Engineeing. Intoduction to obotics Sample Poblems and Solutions fo the Mid-em Exam Figue shows a obotic vehicle having two poweed

More information

Fast quality-guided flood-fill phase unwrapping algorithm for three-dimensional fringe pattern profilometry

Fast quality-guided flood-fill phase unwrapping algorithm for three-dimensional fringe pattern profilometry Univesity of Wollongong Reseach Online Faculty of Infomatics - Papes (Achive) Faculty of Engineeing and Infomation Sciences 2010 Fast quality-guided flood-fill phase unwapping algoithm fo thee-dimensional

More information

High performance CUDA based CNN image processor

High performance CUDA based CNN image processor High pefomance UDA based NN image pocesso GEORGE VALENTIN STOIA, RADU DOGARU, ELENA RISTINA STOIA Depatment of Applied Electonics and Infomation Engineeing Univesity Politehnica of Buchaest -3, Iuliu Maniu

More information

On Error Estimation in Runge-Kutta Methods

On Error Estimation in Runge-Kutta Methods Leonado Jounal of Sciences ISSN 1583-0233 Issue 18, Januay-June 2011 p. 1-10 On Eo Estimation in Runge-Kutta Methods Ochoche ABRAHAM 1,*, Gbolahan BOLARIN 2 1 Depatment of Infomation Technology, 2 Depatment

More information

Topological Characteristic of Wireless Network

Topological Characteristic of Wireless Network Topological Chaacteistic of Wieless Netwok Its Application to Node Placement Algoithm Husnu Sane Naman 1 Outline Backgound Motivation Papes and Contibutions Fist Pape Second Pape Thid Pape Futue Woks Refeences

More information

Complete Solution to Potential and E-Field of a sphere of radius R and a charge density ρ[r] = CC r 2 and r n

Complete Solution to Potential and E-Field of a sphere of radius R and a charge density ρ[r] = CC r 2 and r n Complete Solution to Potential and E-Field of a sphee of adius R and a chage density ρ[] = CC 2 and n Deive the electic field and electic potential both inside and outside of a sphee of adius R with a

More information

Accurate Diffraction Efficiency Control for Multiplexed Volume Holographic Gratings. Xuliang Han, Gicherl Kim, and Ray T. Chen

Accurate Diffraction Efficiency Control for Multiplexed Volume Holographic Gratings. Xuliang Han, Gicherl Kim, and Ray T. Chen Accuate Diffaction Efficiency Contol fo Multiplexed Volume Hologaphic Gatings Xuliang Han, Gichel Kim, and Ray T. Chen Micoelectonic Reseach Cente Depatment of Electical and Compute Engineeing Univesity

More information

Multi-azimuth Prestack Time Migration for General Anisotropic, Weakly Heterogeneous Media - Field Data Examples

Multi-azimuth Prestack Time Migration for General Anisotropic, Weakly Heterogeneous Media - Field Data Examples Multi-azimuth Pestack Time Migation fo Geneal Anisotopic, Weakly Heteogeneous Media - Field Data Examples S. Beaumont* (EOST/PGS) & W. Söllne (PGS) SUMMARY Multi-azimuth data acquisition has shown benefits

More information

The EigenRumor Algorithm for Ranking Blogs

The EigenRumor Algorithm for Ranking Blogs he EigenRumo Algoithm fo Ranking Blogs Ko Fujimua N Cybe Solutions Laboatoies N Copoation akafumi Inoue N Cybe Solutions Laboatoies N Copoation Masayuki Sugisaki N Resonant Inc. ABSRAC he advent of easy

More information

Query Language #1/3: Relational Algebra Pure, Procedural, and Set-oriented

Query Language #1/3: Relational Algebra Pure, Procedural, and Set-oriented Quey Language #1/3: Relational Algeba Pue, Pocedual, and Set-oiented To expess a quey, we use a set of opeations. Each opeation takes one o moe elations as input paamete (set-oiented). Since each opeation

More information

Introduction to Medical Imaging. Cone-Beam CT. Introduction. Available cone-beam reconstruction methods: Our discussion:

Introduction to Medical Imaging. Cone-Beam CT. Introduction. Available cone-beam reconstruction methods: Our discussion: Intoduction Intoduction to Medical Imaging Cone-Beam CT Klaus Muelle Available cone-beam econstuction methods: exact appoximate Ou discussion: exact (now) appoximate (next) The Radon tansfom and its invese

More information

COLOR EDGE DETECTION IN RGB USING JOINTLY EUCLIDEAN DISTANCE AND VECTOR ANGLE

COLOR EDGE DETECTION IN RGB USING JOINTLY EUCLIDEAN DISTANCE AND VECTOR ANGLE COLOR EDGE DETECTION IN RGB USING JOINTLY EUCLIDEAN DISTANCE AND VECTOR ANGLE Slawo Wesolkowski Systems Design Engineeing Univesity of Wateloo Wateloo (Ont.), Canada, NL 3G s.wesolkowski@ieee.og Ed Jenigan

More information

Mobility Pattern Recognition in Mobile Ad-Hoc Networks

Mobility Pattern Recognition in Mobile Ad-Hoc Networks Mobility Patten Recognition in Mobile Ad-Hoc Netwoks S. M. Mousavi Depatment of Compute Engineeing, Shaif Univesity of Technology sm_mousavi@ce.shaif.edu H. R. Rabiee Depatment of Compute Engineeing, Shaif

More information

An Assessment of the Efficiency of Close-Range Photogrammetry for Developing a Photo-Based Scanning Systeminthe Shams Tabrizi Minaret in Khoy City

An Assessment of the Efficiency of Close-Range Photogrammetry for Developing a Photo-Based Scanning Systeminthe Shams Tabrizi Minaret in Khoy City Austalian Jounal of Basic and Applied Sciences, 5(1): 80-85, 011 ISSN 1991-8178 An Assessment of the Efficiency of Close-Range Photogammety fo Developing a Photo-Based Scanning Systeminthe Shams Tabizi

More information

Gravitational Shift for Beginners

Gravitational Shift for Beginners Gavitational Shift fo Beginnes This pape, which I wote in 26, fomulates the equations fo gavitational shifts fom the elativistic famewok of special elativity. Fist I deive the fomulas fo the gavitational

More information

Performance Optimization in Structured Wireless Sensor Networks

Performance Optimization in Structured Wireless Sensor Networks 5 The Intenational Aab Jounal of Infomation Technology, Vol. 6, o. 5, ovembe 9 Pefomance Optimization in Stuctued Wieless Senso etwoks Amine Moussa and Hoda Maalouf Compute Science Depatment, ote Dame

More information

View Synthesis using Depth Map for 3D Video

View Synthesis using Depth Map for 3D Video View Synthesis using Depth Map fo 3D Video Cheon Lee and Yo-Sung Ho Gwangju Institute of Science and Technology (GIST) 1 Oyong-dong, Buk-gu, Gwangju, 500-712, Republic of Koea E-mail: {leecheon, hoyo}@gist.ac.k

More information

Topic 7 Random Variables and Distribution Functions

Topic 7 Random Variables and Distribution Functions Definition of a Random Vaiable Distibution Functions Popeties of Distibution Functions Topic 7 Random Vaiables and Distibution Functions Distibution Functions 1 / 11 Definition of a Random Vaiable Distibution

More information

Haptic Glove. Chan-Su Lee. Abstract. This is a final report for the DIMACS grant of student-initiated project. I implemented Boundary

Haptic Glove. Chan-Su Lee. Abstract. This is a final report for the DIMACS grant of student-initiated project. I implemented Boundary Physically Accuate Haptic Rendeing of Elastic Object fo a Haptic Glove Chan-Su Lee Abstact This is a final epot fo the DIMACS gant of student-initiated poject. I implemented Bounday Element Method(BEM)

More information

Detection and tracking of ships using a stereo vision system

Detection and tracking of ships using a stereo vision system Scientific Reseach and Essays Vol. 8(7), pp. 288-303, 18 Febuay, 2013 Available online at http://www.academicjounals.og/sre DOI: 10.5897/SRE12.318 ISSN 1992-2248 2013 Academic Jounals Full Length Reseach

More information

Improved Fourier-transform profilometry

Improved Fourier-transform profilometry Impoved Fouie-tansfom pofilomety Xianfu Mao, Wenjing Chen, and Xianyu Su An impoved optical geomety of the pojected-finge pofilomety technique, in which the exit pupil of the pojecting lens and the entance

More information

arxiv: v2 [physics.soc-ph] 30 Nov 2016

arxiv: v2 [physics.soc-ph] 30 Nov 2016 Tanspotation dynamics on coupled netwoks with limited bandwidth Ming Li 1,*, Mao-Bin Hu 1, and Bing-Hong Wang 2, axiv:1607.05382v2 [physics.soc-ph] 30 Nov 2016 1 School of Engineeing Science, Univesity

More information

A Haptic-Assisted Guidance System for Navigating Volumetric Data Sets

A Haptic-Assisted Guidance System for Navigating Volumetric Data Sets A Haptic-Assisted Guidance System fo Navigating Volumetic Data Sets Eick Méndez*, Shunsuke Yoshida, Hauo Noma, Robet W. Lindeman*, Yasuyuki Yanagida, Shinobu Masaki, Kenichi Hosaka (*)Compute Science Depatment,

More information

IP Multicast Simulation in OPNET

IP Multicast Simulation in OPNET IP Multicast Simulation in OPNET Xin Wang, Chien-Ming Yu, Henning Schulzinne Paul A. Stipe Columbia Univesity Reutes Depatment of Compute Science 88 Pakway Dive South New Yok, New Yok Hauppuage, New Yok

More information

Conservation Law of Centrifugal Force and Mechanism of Energy Transfer Caused in Turbomachinery

Conservation Law of Centrifugal Force and Mechanism of Energy Transfer Caused in Turbomachinery Poceedings of the 4th WSEAS Intenational Confeence on luid Mechanics and Aeodynamics, Elounda, Geece, August 1-3, 006 (pp337-34) Consevation Law of Centifugal oce and Mechanism of Enegy Tansfe Caused in

More information