First Order Statistics Classification of Sidescan Sonar Images from the Arctic Sea Ice

Size: px
Start display at page:

Download "First Order Statistics Classification of Sidescan Sonar Images from the Arctic Sea Ice"

Transcription

1 irst Order Statistics Classification of Sidescan Sonar Images from the Arctic Sea Ice S. Rueda, Dr. J. Bell, Dr. C. Capus Ocean Systems Lab. School of Engineering and Physical Sciences Heriot Watt University Riccarton EH14 4AS, Edinburgh, Scotland, UK Abstract Polar regions, especially sensitive to small changes in temperature, play a key role in global climate change. Scientists are interested in evaluating the decline in local ice production. We will attempt to classify automatically irst Year (Y) ice, Multi Year (MY) ice and Deformed ice using Sidescan sonar images of the Arctic ice-shelf. We use 4-bit data (16 grey levels) ground truth images provided by an expert as a starting point of the study. These images have been obtained using a Sidescan sonar pointing upward. Our methods use first order statistics extracted from local areas of the image. The local histogram is fitted to three pdfs (Rayleigh, Log-normal and Gaussian distributions) whose parameters are extracted. A Chi-square test is used to evaluate the quality of the fit. The parameters are then used to classify the regions. The results obtained show that Y and MY ice follow Rayleigh or Log-normal distributions whereas Deformed ice is more like a Gaussian distribution. We have created a new method to classify ice types selecting the best fitting for each region. A classification of three classes (Y, MY and Deformed ice) is achieved with first order statistics. In future work we will investigate the potential of second order statistics to improve the classification. I. INTRODUCTION Changes in the Arctic region are closely related to global climate change. To study the interaction between the Arctic and climate, some expeditions have been undertaken and data has been extracted from the Pole. Our aim is to know the extent and the thickness of the sea ice, because we want to evaluate the decline in local ice production which can be an indication of climate change. As a result, we need to find out how much new ice has been formed, and how much is old. Other features of the ice cover, like Deformed ice regions, will give us more information about how climate change is affecting the Arctic. Sidescan sonar has been used to provide images from underneath the polar cap in the Arctic. The sidescan sonar has been put on a submarine which has travelled underneath the polar ice region creating images of the ice cover. With these images we want to identify environmental impacts on polar ice. As we have several thousand kilometres of image data, we will attempt to do an automated and consistent classification of ice types. We work with low quality images coded with 4- bit data (16 grey levels). We use as a base ground truth images manually classified by an expert on ice types. According to Sear and Wadhams [3] and Wadhams [4], each ice type presents its own characteristic features as well as some statistical properties. We can distinguish: - Multi Year ice (MY): rough surface. It presents hummocks and depressions. - irst Year ice (Y): smooth texture, generally featureless, though often criss-crossed with a pattern of narrow cracks. - Pressure ridges (R): linear sets of broken ice blocks. - Lead: represents open water or a zone recently refrozen. - Randomly Deformed ice (D): matrix with no identifiable structure. As stated by Sear and Wadhams in [3], MY ice, Deformed ice and Ridges are more likely to be juxtaposed, as well Y ice with leads. In the first stage, we redefine Deformed ice as rugged areas, usually with a white overall appearance. They rarely occur as simple bounded shapes. These often represent the infill areas between the more coherent Y and MY ice floes and include the linear ridge features. We will attempt to classify three classes: Y, MY and Deformed ice. We want to find out the statistical properties of each ice type texture to achieve a classification. A texture can be described by a random process. If the first order probability density function is invariant with spatial position, then the random process can be considered first order stationary. That means, that all the pixels in the image have the same probability of having a particular grey level. The first order probability density function of an image is its histogram. The first part of this study consist of explaining the methods that we use to achieve a classification of ice types. Then, in a second part we state the results obtained and finally we give a conclusion. The methods presented here, use histograms and probability density functions (pdf) to evaluate pixels distributions in the regions of interest (ROI). Then, we work with the pdf (because it is independent of the size of the ROI), subject to some minimum size constraint, to create a model of the regions selected. Therefore, we apply curve fitting techniques to the 1

2 pdf obtained in order to produce a minimum parameter model. We use Rayleigh, Log-normal and Gaussian distributions to fit our data. Then, we apply these parameters to classify ice types. The algorithms are tested on ground truth images to see the correspondence with the manually classified data. We present the results in one segmented image displaying the three ice types that we are interested on. II. METHODS A. The histogram and probability density function The histogram gives information about the frequency of occurence of grey levels in an image or region of interest. We have plotted the histograms for each ice type to see if there is a significant difference between them. To make the results independent of the number of pixels, we divide the histograms by the total number of pixels. We then obtain the probability density function of the regions of interest. If we take several regions of each ice type, the pdfs obtained are shown in ig.1. Probability density ig. 1. PD COMPARISON Deformed ice Deformed ice Y ice Y ice Y ice MY ice MY ice Gray level PD comparison of ice types We see that we can easily distinguish Y, MY and Deformed ice using the pdf. To produce a minimum parameter model of texture, standard curve fitting techniques can be applied to obtain an approximate fit to the curve of the image pdf. Then only the parameters required to describe the curves need to be stored to create a model to fit the texture. Y ice and MY ice seems to follow a Rayleigh or Log-normal distribution. Deformed is more like a Gaussian distribution. B. Distribution itting The sidescan sonar images that we are processing are the result of an acoustic field scattered from a random ice surface. According to Bell [1], the acoustic field is complex and follows Gaussian distributions when the number of elementary waves scattered tends to infinity. In addition, the amplitude of the field will have a Rayleigh distribution if its real and imaginary parts are uncorrelated and independent. This is the case for sidescan sonar images which are expected to have Rayleigh statistics except in the region of the image where the beam sent by the sonar was normal to the surface. The rougher the surface is, the less Rayleigh distribution it has. This is why the pdfs of Y and MY ice seem to follow Rayleigh distributions and Deformed ice looks more like a Gaussian distribution. This is due to the fact that deformed regions are very rough. 1) The Rayleigh distribution: the Rayleigh pdf is defined (1) in [2] by: where represents the grey level magnitude and is the scale parameter which is equal to the mean-square value of the individual amplitudes :. We have introduced an offset parameter in the formula to allow translations of the Rayleigh pdf in the axis. Therefore, we transform a one parameter distribution into a two-parameter distribution. This means that we can fit better our pdfs because they do not start always at. If we call the offset, the formula becomes:!#" $ %'& )(" -+*), /.) (2) We will try to apply fitting techniques to our data to describe the curves only with the parameters associated to the distributions. We use a least square error technique to find the best estimation of the data. This method consists of calculating the squared error, which is the sum of the squared differences between each sample and the expected value as predicted by the function. The least square error has the general expression: 789<?>A@ ;:! CBED (3) where! are the values of the original data andd are the values of the curve obtained by fitting the data. The results obtained with a Rayleigh fitting are shown in ig. 2. Probability Density unction RAYLEIGH ITTING Grey level ig. 2. The Rayleigh distribution fitting Apparently we can fit a Rayleigh in Y and MY ice regions, but for Deformed ice, the fitting does not look good. 2

3 M Y Y Z Z We cannot distinguish MY and Deformed ice just with the plots. We need to analyze the parameters of these distributions fitted to see if they can help to distinguish MY and Deformed ice. To separate the distributions we have used the parameter which is related HG - to the Rayleigh scale as:. The parameters and offset obtained are shown in Table I. TABLE I RAYLEIGH PARAMETERS b offset Y MY D The rougher the region of ice selected, the higher the parameter is. The parameters are different for each ice type selected. We will see in the part of the Results how good is the classification using this method. 2) The Log-normal distribution: it is a two-parameter distribution described by its mean I and its standard deviation J. The formula used to fit the data is: DK! $ LI J J N PO.3QRS.2TVU)2 (4) XW If we apply curve fitting techniques to our data using Lognormal distributions, we get ig. 3. Probability Density unction LOG NORMAL ITTING Grey level ig. 3. The Log-normal distribution fitting As before, the plots of MY and Deformed ice are located in the same ranks of values. We need to have a look at the parameters to see if they are sufficient to distinguish between ice types. They are shown in Table II. As with Rayleigh distributions, the parameters are higher in MY and Deformed ice than in Y ice. This is also due to the roughness of the surface. A classification of three ice types can be achieved considering both parameters. TABLE II LOG-NORMAL PARAMETERS Y MY D ) The Gaussian distribution: it is a bell shaped model defined by two parameters: the mean I and the standard deviation J. This kind of distribution follows the formula: DK! $ LI J J N M #O.)VU)2 (5) XW If we use this distribution to fit our data we get the curves in ig. 4. Probability density function GAUSS ITTING Grey level ig. 4. The Gaussian distribution fitting Only Deformed ice is well fitted by Gaussian distributions. The parameters I and J obtained for all the fittings presented are shown in Table III. TABLE III GAUSSIAN PARAMETERS Y MY D As before, the rougher the surface is, the higher the mean is. But we know from inspection of ig. 4. that this distribution does not fit Y and MY ice well, only Deformed ice. We will see in section III that a good classification cannot be achieved using Gaussian distributions. 3

4 ^ 4) Comparison between fitting techniques: we have used a Chi-square ([ ) test to determine the consistency of the fitting. It compares the pdf data with the fitted distribution for each method applied. The formula used is: [ <\ $] B_^ (6) ] where are the values of the pdf of our data, which represents the probability of pixels within the region selected having a grey level ` ^, and is the value of the fitted distribution at grey level`. Applying this method for the same representative regions, we can see how good are our fittings. We have selected some characteristic regions for the three ice types considered, the [ values obtained for all the distributions used are shown in Table IV. TABLE IV CHI SQUARE PARAMETERS Rayleigh Log-normal Gaussian Y acbdadfehgjikl acba)mjkjmhgjik)n MY D If we analyze these results, we can notice that for Y ice, a good fitting is found with Rayleigh and Log-normal in most of the cases. The best fitting is found using a Log-normal distribution because it has the lowest values for the regions of interest selected in this case. The Gaussian distribution does not fit Y ice well, because the[ value is very big. or MY ice, all three distributions seems to fit the data well. Overall, the best fitting is again obtained when we use the Log-normal distribution to fit the data. But the Rayleigh distribution provides the best fit for one case only. The Gaussian distributions fit MY ice well but they are worse than the other two distributions because the [ value is always higher. or Deformed ice, Gaussian distribution seems to fit the data better because the values represented are lower than the others. or the other two distributions, the Rayleigh distribution seems more consistent than the Log-normal distribution. Y, MY and Deformed. 1) Classification using the fitted Rayleigh, Log-normal and Gaussian parameters: firstly, we define some training patterns representative of each class (Y, MY and Deformed ice) taken from various images. Then, we fit the distributions to them and get the associated parameters. Secondly, we form a vector of targets which provides correspondence between the parameters of the training patterns and the classes they belong to. Thirdly, we cut the image selected into regions of interest (ROIs), with or without overlapping, of a user-defined size and apply the fitting techniques to obtain the distribution parameters of the regions. After analyzing the statistical properties of regions with different sizes, we can state that the minimum size of the regions that we can use without losing the statistical properties is 25 x 25 pixels. inally, a linear discriminant classifier is used to compare the parameters obtained for each region with those of the training patterns to assign a class to each of the image regions. When we get the class of each region we form a new image putting each class in one colour. We have defined Y ice in black, MY ice in white and Deformed ice in grey. 2) Classification using the best fitted parameters: we have seen that Y and MY ice were well fitted with Log-normal and Rayleigh distributions because they presented the lowest [ values for these ice types. Nevertheless, Deformed ice was best fitted by Gaussian distributions. Consequently, it would be interesting to use the [ value to perform our classification, and get the best results than we can. Each region will be classified by the method which best fits its pdf. Therefore, we have stated a new method (cf. ig. 5) which selects for each ROI considered in the image, the best fitted distribution comparing the [ values for the Rayleigh, Lognormal and Gaussian distributions. In other words, as before, we divide the image into regions and we classify each of them independently. We calculate the Rayleigh, Log-normal and Gaussian parameters and the[ value for each ROI. We assign the method that we need to use for each region considering the lowest [ value. With this method we achieve the best classification for each region considered. These results agree with our observations from the plots, which are that Y and MY ice are best fitted by a Log-normal or a Rayleigh distribution but Deformed ice is best fitted by a Gaussian distribution. C. Classification with irst Order Statistics Using the parameters obtained with the methods presented above, we can achieve a segmentation of the three ice types: ig. 5. Schematic of classification process 4

5 III. RESULTS We use the image ig. 6. because it has all three classes and we have a full ground truth classmap as shown in ig. 7. We have applied this methods to other images and the results of the segmentation found are also good. The image that we will attempt to classify has 179 x 2852 pixels and is shown in ig. 6. ig. 9. Classification using Rayleigh distributions (non overlapped regions of size 25x25 parametero) It seems that classifying only with the parameter gives a more accurate classification, and the result is closer to the classes defined in the ground truth image. ig. 6. Original image Y ice, MY ice and Deformed ice (D) The ground truth image that we use is shown in ig. 7. In order to follow class boundaries of the original image in a smoother way, we consider overlapped regions of size 25 x 25, shifting one pixel at a time. We get the result in ig. 1. ig. 7. Ground Truth Image Y ice, MY ice and Deformed ice (D) ig. 1. Classification using Rayleigh distributions (overlapping regions of size 25x25) A. Classification using Rayleigh parameters If we use the parameters and offset to achieve a classification, for non-overlapped regions with size 25 x 25, we get the result presented in ig. 8. The result is very good but the algorithm with overlapping is comparatively very slow because we need to perform 25 x 25 times as many calculations. We extract the boundary contours for each class and plot over the original image to see how good the classification is. The result is shown in ig. 11. ig. 8. Classification using Rayleigh distributions (non overlapped regions of size 25x25 parameters o and offset) If we use only to classify, the result appears in ig. 9. ig. 11. Contour result overlapped on the original image (MY ice: white contour Deformed ice: red contour Y ice: others) 5

6 B. Classification using Log-normal parameters As before, using non-overlapped regions of size 25 x 25, we get ig. 12. This offers the best accuracy/time trade-off, so we did not continue with overlapped regions. ig. 14. Classification using best fitted distributions (non overlapped regions of size 25x25) ig. 12. Classification using Log-normal distributions (non overlapped regions of size 25x25) C. Classification using Gaussian parameters We found that if classes were confused with a large region size, the results cannot be improved with reduced region sizes. With Gaussian parameters and dividing the image into nonoverlapped regions with size 5 x 5, we obtain ig. 13. IV. CONCLUSION We presented a classification of ice types using first order statistics and curve fitting techniques. The results obtained show that Y and MY ice are well fitted by Rayleigh or Log-normal distributions whereas Deformed ice tends to follow a Gaussian distribution. Applying a linear classifier with these three distribution types, we got good results if we used Rayleigh or Log-normal distributions but with Gaussian distributions confusion appears between MY and Deformed ice. or this reason we proposed another method to select the best classification for each region of the image using a [ parameter. This gave us the best results with curve fitting techniques. We found that the minimum size ROIs that we could use to conserve the statistical properties was 25x25 pixels. If the regions of the image are overlapped when we do the classification, the results more accurately follow the class boundaries of the original image. ig. 13. Classification using Gaussian distributions (non overlapped regions of size 5x5) As we can see there is a confusion between MY and Deformed ice. Therefore we do not need to classify this with smaller regions because the result will not be better. D. Classification using best fitted parameters Using the method presented above to select the distribution which best fits each region, without overlapping and for regions with size of 25 x 25, we achieve the classification shown in ig. 14. ACKNOWLEDGMENTS The sidescan sonar images used in this study were provided by The Scottish Association for Marine Science. In particular we wish to thank Martin Doble for providing the ground truth images that helped us in classification of the ice types. This work has been partially supported by the 5th ramework Program of research of the European Community through the AMASON (EVK3-CT-21-59) project and by the Conseil Régional de Bretagne through the Ulysse fellowship. REERENCES [1] J. BELL, A model for the Simulation of Sidescan Sonar, PhD Thesis, Heriot-Watt University, 1995 [2] G.R. ELSTON, Sonar Simulation a Visualisation Using an Acoustic Pseudospectral time-domain Model, PhD Thesis, Heriot-Watt University, 23 [3] C.B SEAR and P. WADHAMS, Statistical properties of Arctic sea ice morphology derived from sidescan sonar images, Progress in Oceanography, Vol. 29, No. 2, pp , 1992 [4] P. WADHAMS, The underside of Arctic sea ice imaged by sidescan sonar, Nature Vol. 333, No. 6169, pp ,

SAR SURFACE ICE COVER DISCRIMINATION USING DISTRIBUTION MATCHING

SAR SURFACE ICE COVER DISCRIMINATION USING DISTRIBUTION MATCHING SAR SURFACE ICE COVER DISCRIMINATION USING DISTRIBUTION MATCHING Rashpal S. Gill Danish Meteorological Institute, Ice charting and Remote Sensing Division, Lyngbyvej 100, DK 2100, Copenhagen Ø, Denmark.Tel.

More information

Sonar Image Compression

Sonar Image Compression Sonar Image Compression Laurie Linnett Stuart Clarke Dept. of Computing and Electrical Engineering HERIOT-WATT UNIVERSITY EDINBURGH Sonar Image Compression, slide - 1 Sidescan Sonar @ Sonar Image Compression,

More information

Region-based Segmentation

Region-based Segmentation Region-based Segmentation Image Segmentation Group similar components (such as, pixels in an image, image frames in a video) to obtain a compact representation. Applications: Finding tumors, veins, etc.

More information

Tutorial: Using Tina Vision s Quantitative Pattern Recognition Tool.

Tutorial: Using Tina Vision s Quantitative Pattern Recognition Tool. Tina Memo No. 2014-004 Internal Report Tutorial: Using Tina Vision s Quantitative Pattern Recognition Tool. P.D.Tar. Last updated 07 / 06 / 2014 ISBE, Medical School, University of Manchester, Stopford

More information

Artifacts and Textured Region Detection

Artifacts and Textured Region Detection Artifacts and Textured Region Detection 1 Vishal Bangard ECE 738 - Spring 2003 I. INTRODUCTION A lot of transformations, when applied to images, lead to the development of various artifacts in them. In

More information

(Refer Slide Time: 0:51)

(Refer Slide Time: 0:51) Introduction to Remote Sensing Dr. Arun K Saraf Department of Earth Sciences Indian Institute of Technology Roorkee Lecture 16 Image Classification Techniques Hello everyone welcome to 16th lecture in

More information

CT Reconstruction with Good-Orientation and Layer Separation for Multilayer Objects

CT Reconstruction with Good-Orientation and Layer Separation for Multilayer Objects 17th World Conference on Nondestructive Testing, 25-28 Oct 2008, Shanghai, China CT Reconstruction with Good-Orientation and Layer Separation for Multilayer Objects Tong LIU 1, Brian Stephan WONG 2, Tai

More information

09/11/2017. Morphological image processing. Morphological image processing. Morphological image processing. Morphological image processing (binary)

09/11/2017. Morphological image processing. Morphological image processing. Morphological image processing. Morphological image processing (binary) Towards image analysis Goal: Describe the contents of an image, distinguishing meaningful information from irrelevant one. Perform suitable transformations of images so as to make explicit particular shape

More information

Lab 9. Julia Janicki. Introduction

Lab 9. Julia Janicki. Introduction Lab 9 Julia Janicki Introduction My goal for this project is to map a general land cover in the area of Alexandria in Egypt using supervised classification, specifically the Maximum Likelihood and Support

More information

Semi-Automatic Transcription Tool for Ancient Manuscripts

Semi-Automatic Transcription Tool for Ancient Manuscripts The Venice Atlas A Digital Humanities atlas project by DH101 EPFL Students Semi-Automatic Transcription Tool for Ancient Manuscripts In this article, we investigate various techniques from the fields of

More information

Emerge Workflow CE8 SAMPLE IMAGE. Simon Voisey July 2008

Emerge Workflow CE8 SAMPLE IMAGE. Simon Voisey July 2008 Emerge Workflow SAMPLE IMAGE CE8 Simon Voisey July 2008 Introduction The document provides a step-by-step guide for producing Emerge predicted petrophysical volumes based on log data of the same type.

More information

Learning and Inferring Depth from Monocular Images. Jiyan Pan April 1, 2009

Learning and Inferring Depth from Monocular Images. Jiyan Pan April 1, 2009 Learning and Inferring Depth from Monocular Images Jiyan Pan April 1, 2009 Traditional ways of inferring depth Binocular disparity Structure from motion Defocus Given a single monocular image, how to infer

More information

6. Object Identification L AK S H M O U. E D U

6. Object Identification L AK S H M O U. E D U 6. Object Identification L AK S H M AN @ O U. E D U Objects Information extracted from spatial grids often need to be associated with objects not just an individual pixel Group of pixels that form a real-world

More information

Scattering of Acoustic Waves from Ocean Boundaries

Scattering of Acoustic Waves from Ocean Boundaries DISTRIBUTION STATEMENT A. Approved for public release; distribution is unlimited. Scattering of Acoustic Waves from Ocean Boundaries Marcia J. Isakson Applied Research Laboratories The University of Texas

More information

Pattern recognition. Classification/Clustering GW Chapter 12 (some concepts) Textures

Pattern recognition. Classification/Clustering GW Chapter 12 (some concepts) Textures Pattern recognition Classification/Clustering GW Chapter 12 (some concepts) Textures Patterns and pattern classes Pattern: arrangement of descriptors Descriptors: features Patten class: family of patterns

More information

Comparison of HK and SC curvature description methods

Comparison of HK and SC curvature description methods Comparison of and SC curvature description methods H. Cantzler and R. B. Fisher Machine Vision Unit, Institute for Perception, Action and Behaviour, Division of Informatics, University of Edinburgh, Edinburgh,

More information

MRI Brain Image Segmentation Using an AM-FM Model

MRI Brain Image Segmentation Using an AM-FM Model MRI Brain Image Segmentation Using an AM-FM Model Marios S. Pattichis', Helen Petropoulos2, and William M. Brooks2 1 Department of Electrical and Computer Engineering, The University of New Mexico, Albuquerque,

More information

AUTOMATIC RECTIFICATION OF SIDE-SCAN SONAR IMAGES

AUTOMATIC RECTIFICATION OF SIDE-SCAN SONAR IMAGES Proceedings of the International Conference Underwater Acoustic Measurements: Technologies &Results Heraklion, Crete, Greece, 28 th June 1 st July 2005 AUTOMATIC RECTIFICATION OF SIDE-SCAN SONAR IMAGES

More information

Investigation of Directional Filter on Kube-Pentland s 3D Surface Reflectance Model using Photometric Stereo

Investigation of Directional Filter on Kube-Pentland s 3D Surface Reflectance Model using Photometric Stereo Investigation of Directional Filter on Kube-Pentland s 3D Surface Reflectance Model using Photometric Stereo Jiahua Wu Silsoe Research Institute Wrest Park, Silsoe Beds, MK45 4HS United Kingdom jerry.wu@bbsrc.ac.uk

More information

CoE4TN4 Image Processing

CoE4TN4 Image Processing CoE4TN4 Image Processing Chapter 11 Image Representation & Description Image Representation & Description After an image is segmented into regions, the regions are represented and described in a form suitable

More information

CLASSIFICATION OF BOUNDARY AND REGION SHAPES USING HU-MOMENT INVARIANTS

CLASSIFICATION OF BOUNDARY AND REGION SHAPES USING HU-MOMENT INVARIANTS CLASSIFICATION OF BOUNDARY AND REGION SHAPES USING HU-MOMENT INVARIANTS B.Vanajakshi Department of Electronics & Communications Engg. Assoc.prof. Sri Viveka Institute of Technology Vijayawada, India E-mail:

More information

Color Image Segmentation

Color Image Segmentation Color Image Segmentation Yining Deng, B. S. Manjunath and Hyundoo Shin* Department of Electrical and Computer Engineering University of California, Santa Barbara, CA 93106-9560 *Samsung Electronics Inc.

More information

Key Terms. Symbology. Categorical attributes. Style. Layer file

Key Terms. Symbology. Categorical attributes. Style. Layer file Key Terms Symbology Categorical attributes Style Layer file Review Questions POP-RANGE is a string field of the Cities feature class with the following entries: 0-9,999, 10,000-49,999, 50,000-99,000 This

More information

Pattern recognition. Classification/Clustering GW Chapter 12 (some concepts) Textures

Pattern recognition. Classification/Clustering GW Chapter 12 (some concepts) Textures Pattern recognition Classification/Clustering GW Chapter 12 (some concepts) Textures Patterns and pattern classes Pattern: arrangement of descriptors Descriptors: features Patten class: family of patterns

More information

Descriptive Statistics, Standard Deviation and Standard Error

Descriptive Statistics, Standard Deviation and Standard Error AP Biology Calculations: Descriptive Statistics, Standard Deviation and Standard Error SBI4UP The Scientific Method & Experimental Design Scientific method is used to explore observations and answer questions.

More information

Machine learning Pattern recognition. Classification/Clustering GW Chapter 12 (some concepts) Textures

Machine learning Pattern recognition. Classification/Clustering GW Chapter 12 (some concepts) Textures Machine learning Pattern recognition Classification/Clustering GW Chapter 12 (some concepts) Textures Patterns and pattern classes Pattern: arrangement of descriptors Descriptors: features Patten class:

More information

GPU Modeling of Ship Operations in Pack Ice

GPU Modeling of Ship Operations in Pack Ice Modeling of Ship Operations in Pack Ice Claude Daley cdaley@mun.ca Shadi Alawneh Dennis Peters Bruce Quinton Bruce Colbourne ABSTRACT The paper explores the use of an event-mechanics approach to assess

More information

Using Statistical Techniques to Improve the QC Process of Swell Noise Filtering

Using Statistical Techniques to Improve the QC Process of Swell Noise Filtering Using Statistical Techniques to Improve the QC Process of Swell Noise Filtering A. Spanos* (Petroleum Geo-Services) & M. Bekara (PGS - Petroleum Geo- Services) SUMMARY The current approach for the quality

More information

Image Segmentation. Segmentation is the process of partitioning an image into regions

Image Segmentation. Segmentation is the process of partitioning an image into regions Image Segmentation Segmentation is the process of partitioning an image into regions region: group of connected pixels with similar properties properties: gray levels, colors, textures, motion characteristics

More information

Practical Image and Video Processing Using MATLAB

Practical Image and Video Processing Using MATLAB Practical Image and Video Processing Using MATLAB Chapter 18 Feature extraction and representation What will we learn? What is feature extraction and why is it a critical step in most computer vision and

More information

2.5 D BEM modelisation of ground structure interaction

2.5 D BEM modelisation of ground structure interaction paper ID: 510/p.1 2.5 D BEM modelisation of ground structure interaction Philippe JEAN CSTB, 24 rue Joseph Fourier, 38400 Saint Martin d Hères, France, jean@cstb.fr 2.5 D Green functions of continuous

More information

Engineering Problem and Goal

Engineering Problem and Goal Engineering Problem and Goal Engineering Problem: Traditional active contour models can not detect edges or convex regions in noisy images. Engineering Goal: The goal of this project is to design an algorithm

More information

Lecture 6: Texture. Tuesday, Sept 18

Lecture 6: Texture. Tuesday, Sept 18 Lecture 6: Texture Tuesday, Sept 18 Graduate students Problem set 1 extension ideas Chamfer matching Hierarchy of shape prototypes, search over translations Comparisons with Hausdorff distance, L1 on

More information

Ensemble registration: Combining groupwise registration and segmentation

Ensemble registration: Combining groupwise registration and segmentation PURWANI, COOTES, TWINING: ENSEMBLE REGISTRATION 1 Ensemble registration: Combining groupwise registration and segmentation Sri Purwani 1,2 sri.purwani@postgrad.manchester.ac.uk Tim Cootes 1 t.cootes@manchester.ac.uk

More information

Boundary descriptors. Representation REPRESENTATION & DESCRIPTION. Descriptors. Moore boundary tracking

Boundary descriptors. Representation REPRESENTATION & DESCRIPTION. Descriptors. Moore boundary tracking Representation REPRESENTATION & DESCRIPTION After image segmentation the resulting collection of regions is usually represented and described in a form suitable for higher level processing. Most important

More information

Scattering from Rock and Rock Outcrops

Scattering from Rock and Rock Outcrops DISTRIBUTION STATEMENT A. Approved for public release; distribution is unlimited. Scattering from Rock and Rock Outcrops Anthony P. Lyons The Pennsylvania State University Applied Research Laboratory,

More information

Texture. Texture is a description of the spatial arrangement of color or intensities in an image or a selected region of an image.

Texture. Texture is a description of the spatial arrangement of color or intensities in an image or a selected region of an image. Texture Texture is a description of the spatial arrangement of color or intensities in an image or a selected region of an image. Structural approach: a set of texels in some regular or repeated pattern

More information

CITS 4402 Computer Vision

CITS 4402 Computer Vision CITS 4402 Computer Vision A/Prof Ajmal Mian Adj/A/Prof Mehdi Ravanbakhsh, CEO at Mapizy (www.mapizy.com) and InFarm (www.infarm.io) Lecture 02 Binary Image Analysis Objectives Revision of image formation

More information

Effects Of Shadow On Canny Edge Detection through a camera

Effects Of Shadow On Canny Edge Detection through a camera 1523 Effects Of Shadow On Canny Edge Detection through a camera Srajit Mehrotra Shadow causes errors in computer vision as it is difficult to detect objects that are under the influence of shadows. Shadow

More information

A New Algorithm for Shape Detection

A New Algorithm for Shape Detection IOSR Journal of Computer Engineering (IOSR-JCE) e-issn: 2278-0661,p-ISSN: 2278-8727, Volume 19, Issue 3, Ver. I (May.-June. 2017), PP 71-76 www.iosrjournals.org A New Algorithm for Shape Detection Hewa

More information

An Intuitive Explanation of Fourier Theory

An Intuitive Explanation of Fourier Theory An Intuitive Explanation of Fourier Theory Steven Lehar slehar@cns.bu.edu Fourier theory is pretty complicated mathematically. But there are some beautifully simple holistic concepts behind Fourier theory

More information

A Geostatistical and Flow Simulation Study on a Real Training Image

A Geostatistical and Flow Simulation Study on a Real Training Image A Geostatistical and Flow Simulation Study on a Real Training Image Weishan Ren (wren@ualberta.ca) Department of Civil & Environmental Engineering, University of Alberta Abstract A 12 cm by 18 cm slab

More information

ESTIMATION OF DETECTION/CLASSIFICATION PERFORMANCE USING INTERFEROMETRIC SONAR COHERENCE

ESTIMATION OF DETECTION/CLASSIFICATION PERFORMANCE USING INTERFEROMETRIC SONAR COHERENCE ESTIMATION OF DETECTION/CLASSIFICATION PERFORMANCE USING INTERFEROMETRIC SONAR COHERENCE Øivind Midtgaard a, Torstein O. Sæbø a and Roy E. Hansen a a Norwegian Defence Research Establishment (FFI), P.

More information

GENERAL AUTOMATED FLAW DETECTION SCHEME FOR NDE X-RAY IMAGES

GENERAL AUTOMATED FLAW DETECTION SCHEME FOR NDE X-RAY IMAGES GENERAL AUTOMATED FLAW DETECTION SCHEME FOR NDE X-RAY IMAGES Karl W. Ulmer and John P. Basart Center for Nondestructive Evaluation Department of Electrical and Computer Engineering Iowa State University

More information

Examination in Image Processing

Examination in Image Processing Umeå University, TFE Ulrik Söderström 203-03-27 Examination in Image Processing Time for examination: 4.00 20.00 Please try to extend the answers as much as possible. Do not answer in a single sentence.

More information

CLASSIFICATION STRATEGIES FOR POLARIMETRIC SAR SEA ICE DATA

CLASSIFICATION STRATEGIES FOR POLARIMETRIC SAR SEA ICE DATA CLASSIFICATION STRATEGIES FOR POLARIMETRIC SAR SEA ICE DATA Bernd Scheuchl (), Irena Hajnsek (), Ian Cumming () () Department of Electrical and Computer Engineering, University of British Columbia 56 Main

More information

MR IMAGE SEGMENTATION

MR IMAGE SEGMENTATION MR IMAGE SEGMENTATION Prepared by : Monil Shah What is Segmentation? Partitioning a region or regions of interest in images such that each region corresponds to one or more anatomic structures Classification

More information

DAMAGE INSPECTION AND EVALUATION IN THE WHOLE VIEW FIELD USING LASER

DAMAGE INSPECTION AND EVALUATION IN THE WHOLE VIEW FIELD USING LASER DAMAGE INSPECTION AND EVALUATION IN THE WHOLE VIEW FIELD USING LASER A. Kato and T. A. Moe Department of Mechanical Engineering Chubu University Kasugai, Aichi 487-8501, Japan ABSTRACT In this study, we

More information

COSC160: Detection and Classification. Jeremy Bolton, PhD Assistant Teaching Professor

COSC160: Detection and Classification. Jeremy Bolton, PhD Assistant Teaching Professor COSC160: Detection and Classification Jeremy Bolton, PhD Assistant Teaching Professor Outline I. Problem I. Strategies II. Features for training III. Using spatial information? IV. Reducing dimensionality

More information

IDENTIFYING GEOMETRICAL OBJECTS USING IMAGE ANALYSIS

IDENTIFYING GEOMETRICAL OBJECTS USING IMAGE ANALYSIS IDENTIFYING GEOMETRICAL OBJECTS USING IMAGE ANALYSIS Fathi M. O. Hamed and Salma F. Elkofhaifee Department of Statistics Faculty of Science University of Benghazi Benghazi Libya felramly@gmail.com and

More information

EMERGE Workflow CE8R2 SAMPLE IMAGE. Simon Voisey Hampson-Russell London Office

EMERGE Workflow CE8R2 SAMPLE IMAGE. Simon Voisey Hampson-Russell London Office EMERGE Workflow SAMPLE IMAGE CE8R2 Simon Voisey Hampson-Russell London Office August 2008 Introduction The document provides a step-by-step guide for producing Emerge-predicted petrophysical p volumes

More information

Chapter 9 Object Tracking an Overview

Chapter 9 Object Tracking an Overview Chapter 9 Object Tracking an Overview The output of the background subtraction algorithm, described in the previous chapter, is a classification (segmentation) of pixels into foreground pixels (those belonging

More information

CSE/EE-576, Final Project

CSE/EE-576, Final Project 1 CSE/EE-576, Final Project Torso tracking Ke-Yu Chen Introduction Human 3D modeling and reconstruction from 2D sequences has been researcher s interests for years. Torso is the main part of the human

More information

Frequency Distributions

Frequency Distributions Displaying Data Frequency Distributions After collecting data, the first task for a researcher is to organize and summarize the data so that it is possible to get a general overview of the results. Remember,

More information

Topic 6 Representation and Description

Topic 6 Representation and Description Topic 6 Representation and Description Background Segmentation divides the image into regions Each region should be represented and described in a form suitable for further processing/decision-making Representation

More information

Time Stamp Detection and Recognition in Video Frames

Time Stamp Detection and Recognition in Video Frames Time Stamp Detection and Recognition in Video Frames Nongluk Covavisaruch and Chetsada Saengpanit Department of Computer Engineering, Chulalongkorn University, Bangkok 10330, Thailand E-mail: nongluk.c@chula.ac.th

More information

Global Probability of Boundary

Global Probability of Boundary Global Probability of Boundary Learning to Detect Natural Image Boundaries Using Local Brightness, Color, and Texture Cues Martin, Fowlkes, Malik Using Contours to Detect and Localize Junctions in Natural

More information

MULTI-DIMENSIONAL FAST MARCHING APPROACH TO WAVE PROPAGATION IN HETEROGENOUS MULTIPATH ENVIRONMENT

MULTI-DIMENSIONAL FAST MARCHING APPROACH TO WAVE PROPAGATION IN HETEROGENOUS MULTIPATH ENVIRONMENT MULTI-DIMENSIONAL FAST MARCHING APPROACH TO WAVE PROPAGATION IN HETEROGENOUS MULTIPATH ENVIRONMENT Yan Pailhas a & Yvan Petillot a a Ocean Systems Lab, Heriot-Watt University Riccarton Campus, EH14 4AS,

More information

Multimedia Computing: Algorithms, Systems, and Applications: Edge Detection

Multimedia Computing: Algorithms, Systems, and Applications: Edge Detection Multimedia Computing: Algorithms, Systems, and Applications: Edge Detection By Dr. Yu Cao Department of Computer Science The University of Massachusetts Lowell Lowell, MA 01854, USA Part of the slides

More information

Detecting Salient Contours Using Orientation Energy Distribution. Part I: Thresholding Based on. Response Distribution

Detecting Salient Contours Using Orientation Energy Distribution. Part I: Thresholding Based on. Response Distribution Detecting Salient Contours Using Orientation Energy Distribution The Problem: How Does the Visual System Detect Salient Contours? CPSC 636 Slide12, Spring 212 Yoonsuck Choe Co-work with S. Sarma and H.-C.

More information

Automatic Segmentation of Semantic Classes in Raster Map Images

Automatic Segmentation of Semantic Classes in Raster Map Images Automatic Segmentation of Semantic Classes in Raster Map Images Thomas C. Henderson, Trevor Linton, Sergey Potupchik and Andrei Ostanin School of Computing, University of Utah, Salt Lake City, UT 84112

More information

Glasses Detection for Face Recognition Using Bayes Rules

Glasses Detection for Face Recognition Using Bayes Rules Glasses Detection for Face Recognition Using Bayes Rules Zhong Jing, Roberto ariani, Jiankang Wu RWCP ulti-modal Functions KRDL Lab Heng ui Keng Terrace, Kent Ridge Singapore 963 Email: jzhong, rmariani,

More information

Image Segmentation. Selim Aksoy. Bilkent University

Image Segmentation. Selim Aksoy. Bilkent University Image Segmentation Selim Aksoy Department of Computer Engineering Bilkent University saksoy@cs.bilkent.edu.tr Examples of grouping in vision [http://poseidon.csd.auth.gr/lab_research/latest/imgs/s peakdepvidindex_img2.jpg]

More information

Image Segmentation. Selim Aksoy. Bilkent University

Image Segmentation. Selim Aksoy. Bilkent University Image Segmentation Selim Aksoy Department of Computer Engineering Bilkent University saksoy@cs.bilkent.edu.tr Examples of grouping in vision [http://poseidon.csd.auth.gr/lab_research/latest/imgs/s peakdepvidindex_img2.jpg]

More information

Image Restoration and Reconstruction

Image Restoration and Reconstruction Image Restoration and Reconstruction Image restoration Objective process to improve an image, as opposed to the subjective process of image enhancement Enhancement uses heuristics to improve the image

More information

Texture Segmentation Using Gabor Filters

Texture Segmentation Using Gabor Filters TEXTURE SEGMENTATION USING GABOR FILTERS 1 Texture Segmentation Using Gabor Filters Khaled Hammouda Prof. Ed Jernigan University of Waterloo, Ontario, Canada Abstract Texture segmentation is the process

More information

Linear combinations of simple classifiers for the PASCAL challenge

Linear combinations of simple classifiers for the PASCAL challenge Linear combinations of simple classifiers for the PASCAL challenge Nik A. Melchior and David Lee 16 721 Advanced Perception The Robotics Institute Carnegie Mellon University Email: melchior@cmu.edu, dlee1@andrew.cmu.edu

More information

Do It Yourself 2. Representations of polarimetric information

Do It Yourself 2. Representations of polarimetric information Do It Yourself 2 Representations of polarimetric information The objectives of this second Do It Yourself concern the representation of the polarimetric properties of scatterers or media. 1. COLOR CODED

More information

Automatic Machinery Fault Detection and Diagnosis Using Fuzzy Logic

Automatic Machinery Fault Detection and Diagnosis Using Fuzzy Logic Automatic Machinery Fault Detection and Diagnosis Using Fuzzy Logic Chris K. Mechefske Department of Mechanical and Materials Engineering The University of Western Ontario London, Ontario, Canada N6A5B9

More information

Computer Vision & Digital Image Processing. Image segmentation: thresholding

Computer Vision & Digital Image Processing. Image segmentation: thresholding Computer Vision & Digital Image Processing Image Segmentation: Thresholding Dr. D. J. Jackson Lecture 18-1 Image segmentation: thresholding Suppose an image f(y) is composed of several light objects on

More information

Chapter 9 Morphological Image Processing

Chapter 9 Morphological Image Processing Morphological Image Processing Question What is Mathematical Morphology? An (imprecise) Mathematical Answer A mathematical tool for investigating geometric structure in binary and grayscale images. Shape

More information

Indoor Object Recognition of 3D Kinect Dataset with RNNs

Indoor Object Recognition of 3D Kinect Dataset with RNNs Indoor Object Recognition of 3D Kinect Dataset with RNNs Thiraphat Charoensripongsa, Yue Chen, Brian Cheng 1. Introduction Recent work at Stanford in the area of scene understanding has involved using

More information

Applications. Foreground / background segmentation Finding skin-colored regions. Finding the moving objects. Intelligent scissors

Applications. Foreground / background segmentation Finding skin-colored regions. Finding the moving objects. Intelligent scissors Segmentation I Goal Separate image into coherent regions Berkeley segmentation database: http://www.eecs.berkeley.edu/research/projects/cs/vision/grouping/segbench/ Slide by L. Lazebnik Applications Intelligent

More information

Experiments on Experiments, Determining Uncertainty for Complex Experiments using Matlab.

Experiments on Experiments, Determining Uncertainty for Complex Experiments using Matlab. Experiments on Experiments, Determining Uncertainty for Complex Experiments using Matlab. Abstract R. Fithen, PhD Arkansas Tech University Often in engineering labs and experiments students expect result

More information

Locating Salient Object Features

Locating Salient Object Features Locating Salient Object Features K.N.Walker, T.F.Cootes and C.J.Taylor Dept. Medical Biophysics, Manchester University, UK knw@sv1.smb.man.ac.uk Abstract We present a method for locating salient object

More information

Implementation and Comparison of Four Different Boundary Detection Algorithms for Quantitative Ultrasonic Measurements of the Human Carotid Artery

Implementation and Comparison of Four Different Boundary Detection Algorithms for Quantitative Ultrasonic Measurements of the Human Carotid Artery Implementation and Comparison of Four Different Boundary Detection Algorithms for Quantitative Ultrasonic Measurements of the Human Carotid Artery Masters Thesis By Ghassan Hamarneh Rafeef Abu-Gharbieh

More information

The Detection of Faces in Color Images: EE368 Project Report

The Detection of Faces in Color Images: EE368 Project Report The Detection of Faces in Color Images: EE368 Project Report Angela Chau, Ezinne Oji, Jeff Walters Dept. of Electrical Engineering Stanford University Stanford, CA 9435 angichau,ezinne,jwalt@stanford.edu

More information

A Review on Plant Disease Detection using Image Processing

A Review on Plant Disease Detection using Image Processing A Review on Plant Disease Detection using Image Processing Tejashri jadhav 1, Neha Chavan 2, Shital jadhav 3, Vishakha Dubhele 4 1,2,3,4BE Student, Dept. of Electronic & Telecommunication Engineering,

More information

ELEC Dr Reji Mathew Electrical Engineering UNSW

ELEC Dr Reji Mathew Electrical Engineering UNSW ELEC 4622 Dr Reji Mathew Electrical Engineering UNSW Review of Motion Modelling and Estimation Introduction to Motion Modelling & Estimation Forward Motion Backward Motion Block Motion Estimation Motion

More information

Points Lines Connected points X-Y Scatter. X-Y Matrix Star Plot Histogram Box Plot. Bar Group Bar Stacked H-Bar Grouped H-Bar Stacked

Points Lines Connected points X-Y Scatter. X-Y Matrix Star Plot Histogram Box Plot. Bar Group Bar Stacked H-Bar Grouped H-Bar Stacked Plotting Menu: QCExpert Plotting Module graphs offers various tools for visualization of uni- and multivariate data. Settings and options in different types of graphs allow for modifications and customizations

More information

Identifying Layout Classes for Mathematical Symbols Using Layout Context

Identifying Layout Classes for Mathematical Symbols Using Layout Context Rochester Institute of Technology RIT Scholar Works Articles 2009 Identifying Layout Classes for Mathematical Symbols Using Layout Context Ling Ouyang Rochester Institute of Technology Richard Zanibbi

More information

A METHOD TO PREDICT ACCURACY OF LEAST SQUARES SURFACE MATCHING FOR AIRBORNE LASER SCANNING DATA SETS

A METHOD TO PREDICT ACCURACY OF LEAST SQUARES SURFACE MATCHING FOR AIRBORNE LASER SCANNING DATA SETS A METHOD TO PREDICT ACCURACY OF LEAST SQUARES SURFACE MATCHING FOR AIRBORNE LASER SCANNING DATA SETS Robert Pâquet School of Engineering, University of Newcastle Callaghan, NSW 238, Australia (rpaquet@mail.newcastle.edu.au)

More information

Using the Kolmogorov-Smirnov Test for Image Segmentation

Using the Kolmogorov-Smirnov Test for Image Segmentation Using the Kolmogorov-Smirnov Test for Image Segmentation Yong Jae Lee CS395T Computational Statistics Final Project Report May 6th, 2009 I. INTRODUCTION Image segmentation is a fundamental task in computer

More information

V. Zetterberg Amlab Elektronik AB Nettovaegen 11, Jaerfaella, Sweden Phone:

V. Zetterberg Amlab Elektronik AB Nettovaegen 11, Jaerfaella, Sweden Phone: Comparison between whitened generalized cross correlation and adaptive filter for time delay estimation with scattered arrays for passive positioning of moving targets in Baltic Sea shallow waters V. Zetterberg

More information

Quantifying the Dynamic Ocean Surface Using Underwater Radiometric Measurements

Quantifying the Dynamic Ocean Surface Using Underwater Radiometric Measurements DISTRIBUTION STATEMENT A. Approved for public release; distribution is unlimited. Quantifying the Dynamic Ocean Surface Using Underwater Radiometric Measurements Dick K.P. Yue Center for Ocean Engineering

More information

WORCESTER POLYTECHNIC INSTITUTE

WORCESTER POLYTECHNIC INSTITUTE WORCESTER POLYTECHNIC INSTITUTE MECHANICAL ENGINEERING DEPARTMENT Optical Metrology and NDT ME-593L, C 2018 Introduction: Wave Optics January 2018 Wave optics: coherence Temporal coherence Review interference

More information

Robust Shape Retrieval Using Maximum Likelihood Theory

Robust Shape Retrieval Using Maximum Likelihood Theory Robust Shape Retrieval Using Maximum Likelihood Theory Naif Alajlan 1, Paul Fieguth 2, and Mohamed Kamel 1 1 PAMI Lab, E & CE Dept., UW, Waterloo, ON, N2L 3G1, Canada. naif, mkamel@pami.uwaterloo.ca 2

More information

Lecture 8 Object Descriptors

Lecture 8 Object Descriptors Lecture 8 Object Descriptors Azadeh Fakhrzadeh Centre for Image Analysis Swedish University of Agricultural Sciences Uppsala University 2 Reading instructions Chapter 11.1 11.4 in G-W Azadeh Fakhrzadeh

More information

VC 11/12 T14 Visual Feature Extraction

VC 11/12 T14 Visual Feature Extraction VC 11/12 T14 Visual Feature Extraction Mestrado em Ciência de Computadores Mestrado Integrado em Engenharia de Redes e Sistemas Informáticos Miguel Tavares Coimbra Outline Feature Vectors Colour Texture

More information

3 Feature Selection & Feature Extraction

3 Feature Selection & Feature Extraction 3 Feature Selection & Feature Extraction Overview: 3.1 Introduction 3.2 Feature Extraction 3.3 Feature Selection 3.3.1 Max-Dependency, Max-Relevance, Min-Redundancy 3.3.2 Relevance Filter 3.3.3 Redundancy

More information

Valleys Deep and Mountains High

Valleys Deep and Mountains High Valleys Deep and Mountains High Purpose With this activity, you will learn and simulate how altitudes and the height of the Earth s surface can be measured from space. Later, you will explore different

More information

Semantic Context Forests for Learning- Based Knee Cartilage Segmentation in 3D MR Images

Semantic Context Forests for Learning- Based Knee Cartilage Segmentation in 3D MR Images Semantic Context Forests for Learning- Based Knee Cartilage Segmentation in 3D MR Images MICCAI 2013: Workshop on Medical Computer Vision Authors: Quan Wang, Dijia Wu, Le Lu, Meizhu Liu, Kim L. Boyer,

More information

An Object Detection System using Image Reconstruction with PCA

An Object Detection System using Image Reconstruction with PCA An Object Detection System using Image Reconstruction with PCA Luis Malagón-Borja and Olac Fuentes Instituto Nacional de Astrofísica Óptica y Electrónica, Puebla, 72840 Mexico jmb@ccc.inaoep.mx, fuentes@inaoep.mx

More information

An Introduction to Content Based Image Retrieval

An Introduction to Content Based Image Retrieval CHAPTER -1 An Introduction to Content Based Image Retrieval 1.1 Introduction With the advancement in internet and multimedia technologies, a huge amount of multimedia data in the form of audio, video and

More information

CHAPTER 6 DETECTION OF MASS USING NOVEL SEGMENTATION, GLCM AND NEURAL NETWORKS

CHAPTER 6 DETECTION OF MASS USING NOVEL SEGMENTATION, GLCM AND NEURAL NETWORKS 130 CHAPTER 6 DETECTION OF MASS USING NOVEL SEGMENTATION, GLCM AND NEURAL NETWORKS A mass is defined as a space-occupying lesion seen in more than one projection and it is described by its shapes and margin

More information

Evaluation of Moving Object Tracking Techniques for Video Surveillance Applications

Evaluation of Moving Object Tracking Techniques for Video Surveillance Applications International Journal of Current Engineering and Technology E-ISSN 2277 4106, P-ISSN 2347 5161 2015INPRESSCO, All Rights Reserved Available at http://inpressco.com/category/ijcet Research Article Evaluation

More information

A Model Based Approach to Mine Detection and Classification in Sidescan Sonar

A Model Based Approach to Mine Detection and Classification in Sidescan Sonar A Model Based Approach to Mine Detection and Classification in Sidescan Sonar S.Reed, Ocean Systems Lab, Heriot-Watt University Y.Petillot, Ocean Systems Lab, Heriot-Watt University J.Bell, Ocean Systems

More information

CoE4TN4 Image Processing. Chapter 5 Image Restoration and Reconstruction

CoE4TN4 Image Processing. Chapter 5 Image Restoration and Reconstruction CoE4TN4 Image Processing Chapter 5 Image Restoration and Reconstruction Image Restoration Similar to image enhancement, the ultimate goal of restoration techniques is to improve an image Restoration: a

More information

Chapter 24. Creating Surfaces for Displaying and Reporting Data

Chapter 24. Creating Surfaces for Displaying and Reporting Data Chapter 24. Creating Surfaces for Displaying and Reporting Data FLUENT allows you to select portions of the domain to be used for visualizing the flow field. The domain portions are called surfaces, and

More information

Impact of 3D Laser Data Resolution and Accuracy on Pipeline Dents Strain Analysis

Impact of 3D Laser Data Resolution and Accuracy on Pipeline Dents Strain Analysis More Info at Open Access Database www.ndt.net/?id=15137 Impact of 3D Laser Data Resolution and Accuracy on Pipeline Dents Strain Analysis Jean-Simon Fraser, Pierre-Hugues Allard Creaform, 5825 rue St-Georges,

More information