TEXT DETECTION SYSTEM FOR THE BLIND

Size: px
Start display at page:

Download "TEXT DETECTION SYSTEM FOR THE BLIND"

Transcription

1 TEXT DETECTION SYSTEM FOR THE BLIND Marcin Pazio, Maciej Niedźwiecki, Ryszard Kowalik, Jacek Lebiedź Faculty of Electronics, Telecommunications and Computer Science Department of Automatic Control, Gdańsk University of Technology Narutowicza 11/12, Gdańsk, Poland Tel: , Fax: s: {mapa, maciekn, kowalik, jacekl}@eti.pg.gda.pl ABSTRACT The system capable of localizing and reading aloud text embedded in natural scene images can be very helpful for blind and visually impaired persons - providing information useful in everyday life, it increases their confidence and autonomy. Even though the currently available optical character recognition (OCR) programs are fast and accurate, most of them fail to recognize text embedded in natural scene images. The goal of the algorithm described in this paper is to localize text-like image regions and pre-process them in a way that will make OCR work more reliably. The approach described in the paper is based on color image segmentation and segment shape analysis. Preliminary tests have shown that the proposed algorithm offers satisfactory detection rate and is pretty robust to typical text distortions, such as slant, tilt and bend. 1. INTRODUCTION Advances in modern information technology and digital signal processing allow one to design devices and computer applications that help visually impaired persons. With the increasing life expectancy the demand for such equipment steadily grows. The GPS-based navigation systems for blind pedestrians, such as the system recently developed in Gdańsk [1], [2], provide support to blind persons when they go outside without assistance. Experiments performed in urban environments by blind volunteers show that, in addition to providing path planning and on-line guidance, such portable navigation devices can play a very important role in exploring the surrounding environment [3]. When combined with the electronic map and its embedded city database, the navigation unit allows blind users to learn about their whereabouts, e.g. to get information about the surrounding shops, service points, bus and tram stops, offices and institutions. It turns out that such information support aspect of navigation systems is very much appreciated by blind persons. Portable text reading devices are useful extensions of navigation systems as they can provide valuable pieces of information, usually not contained in the electronic databases. Street information boards, shop signboards, text signs besides the office entrances and traffic signs with a text content (such as stop signs, for example) are often the source of important and/or helpful text messages that can be acquired and read aloud to a blind person by a camera-based text reading system. Since the commercially available OCR systems are designed to process typical clean documents, scanned under good illumination conditions, they usually fail to read text messages embedded in natural scene images, where the text background is textured or heavily cluttered, text fonts may differ in size, orientation and alignment, lighting conditions are uneven etc. Therefore, to successfully use the OCR software one should localize text areas first [4]. Text localization in natural scenes has been a subject of many recent studies [4], [5], [6], [7], [8], [9], [10], [11], [12]. The problem is usually solved in three steps as follows. First, the image is searched for text-like areas. The most commonly used approaches involve: (1) Color-based image segmentation, followed by shape analysis of feasible segments [6], [10] (2) Local analysis of the image edge maps [6], [10] (3) Application of morphological operations (such as top-hat transformations) [6], [7] (4) Local analysis of the spatial frequency image structure (using Gabor filters, for example) [5], [9] The classifiers, usually combining many different local features, range from very simple OR-type combiners, through time-efficient cascade detection systems, to fairly complex decision networks optimized using the machine learning methods [12]. In the second step, the detected text areas are combined into larger blocks or strings. This task is usually performed using the connected component analysis based on different similarity criteria, such as similarity of size, texture and color, spatial closeness, alignment etc. Finally, the bounding boxes of the text strings are generated and the annotated image is passed on to the OCR software for higher-level processing. 2. SUMMARY OF THE TEXT-DETECTION ALGORITHM One of our main concerns at the time of starting the text detection project was how to guarantee sufficient degree of robustness of the text localization algorithm to typical text distortions, such as slant, tilt and bend. One should realize that in the case considered text distortions are practically unavoidable. First, images of natural scenes are projections from the 3-D space and therefore many text elements are 2007 EURASIP 272

2 subject to perspective distortions. Second, blind users are not very good at keeping the camera horizontally aligned, which introduces some additional degree of slantness, even if perspective distortions are absent. Therefore, even though we have focused on localization of horizontally aligned text strings (which constitute majority of text messages embedded in natural scene images) we had to assume, right from the beginning, that our system should cope favorably with high degree of text slantness, up to ±45 o or so. The proposed text detection algorithm combines classical approaches, mentioned in Section 1, with some new techniques. It consists of three steps: color image segmentation, featurebased segment filtration and hierarchical segment clusterization. All steps will be described in some detail in the subsequent sections. 3. COLOR IMAGE SEGMENTATION Most of text messages encountered in the outdoor environment are written with colored letters, on a background that assures adequate contrast. Usually the letters have the same color within the text. Therefore information about colors may be very useful in spotting text regions. The segmentation method we use tries to preserve as much of the color content of the original image as possible. The image segmentation process is, in general, a time-consuming operation, especially when it is required to arrive at a reasonable number of properly shaped segments. For the OCR to work properly the shapes of segments should preserve the shapes of letters. The proposed solution is a compromise between the processing speed and accuracy. The segmentation algorithm was designed so as to preserve the basic properties of letters, such as their height, width, general shape and color. The shape of the letter-like segments can be later corrected at the second stage of processing called resegmentation. The applied segmentation algorithm is of the region growth type. A brief summary of the segmentation procedure is given below. (1) Convert original image to the Lab color model. (2) Create the edge image based on the L component of the original image. (3) Equalize histograms of the L, a, and b components using 5 bits coding. (4) Create the 3-D Lab histogram of the image, neglecting all edge pixels. (5) Find the color corresponding to the maximum of the histogram. (6) Take any pixel corresponding to the histogram maximum and regard it as a seed pixel. Grow the segment around the seed pixel; when finished correct (decrease) the corresponding histogram values. (7) If all values of the corrected histogram are zero stop, otherwise go to step 5. In the first step the image is converted to the Lab colorspace. This model of color coding allows one to choose between lightness and color as properties of the processed image. The second step is used to create an auxiliary edge image based on the L component of the Lab model. Histogram equalization, performed in the third step, significantly reduces the 3-D Lab histogram size, while preserving the image readability. A typical 24-bit RGB color image of a natural scene contains less than different colors from the palette of the 2 83 = possible ones. The 3 5 bit coding is sufficient to preserve important image details; at the same time it reduces the size of the histogram, created in the forth step, to values. The result of the fourth step is the 3-D histogram matrix. To avoid creation of segments that consist of edge pixels, all such pixels are excluded from the histogram. The fifth step is used to find the maximum of the histogram. The L, a and b coordinates of this maximum determine the color of seed pixels. The iterative procedure performed in the sixth step grows segments around the seed pixels. The weighted distance in the Lab space is used as the region homogeneity measure, namely a new pixel is added to the existing segment if δ = w 1 L a L s + w 2 ( a a a s + b a b s ) δ max where L a, a a, b a and L s, a s, b s denote the Lab color coordinates of the analyzed pixel and the seed pixel, respectively. The weights w 1 and w 2 (trimmed experimentally) were introduced to make the analysis less sensitive to effects caused by uneven lighting. Each pixel added to the existing segment modifies (decreases) the corresponding histogram value, except for the situation where the pixel is placed on the edge. The last, seventh step is used to check the stopping condition; when the stopping criterion is fulfilled the segmentation process comes to an end. 4. FEATURE-BASED SEGMENT FILTRATION The goal of the feature based segment filtration is to reduce the number of candidate segments further considered. Several geometrical properties of segments are analyzed. Segments that do not have the letter-like shape are marked as not letters. The features we use for such a preliminary screening are: the absolute segment size the relative segment height the height to width ratio the bounding rectangle fill ratio versus the height to width ratio 4.1 Absolute segment size The absolute segment size test allows us to eliminate very small and very large segments. Small segments, containing less than 10 pixels, are usually segmentation artifacts, or they correspond to characters that are too small to be properly recognized by OCR. Large segments, with size comparable to the image size, correspond usually to non-text objects (e.g architectonic elements or the sky). 4.2 Relative segment height The relative segment height test amounts to checking whether the relative segment height α obeys the following inequality α = S w α max 2007 EURASIP 273

3 Fig. 1. The bounding rectangle fill ratio versus the height to width ratio test where S denotes the segment area (in pixels), w denotes the width of the bounding rectangle and α max is the experimentally determined threshold. This test allows us to eliminate small squares and vertical elements that are too small to be considered as letters or digits. 4.3 Height to width ratio The height to width ratio test checks whether the height to width ratio β obeys β = h w β max where h and w denote the height and width of the bounding rectangle, respectively. This test allows us to eliminate segments that are too narrow and too high. In particular, it eliminates high, vertical segments (usually corresponding to architectonic details) which can be easily confused with the letter I. 4.4 Bounding rectangle fill ratio versus the height to width ratio The bounding rectangle fill ratio versus the height to width ratio test is based on the intuitive assumption, that only the vertically positioned I -shaped alphanumerical signs are close in its shape to rectangles. Therefore most of the vertical rectangle-like shapes may be eliminated. In real images this relation is less trivial. Figure 1 shows typical values of the two ratios observed for the letter-shaped segments (each point corresponds to one segment). The segments with relatively high values of the fill ratio correspond to fragments of letters, usually obtained as a result of oversegmentation. In most cases oversegmentation is caused by uneven lightning conditions. The test is a graphical one - all segments that fall into the shaded areas in Figure 1 are rejected as not letters. Fig. 2. Formation of elementary text clusters. 5. CONNECTED COMPONENT ANALYSIS In order to localize large text-like image areas we use the hierarchical clustering procedure. First, the analysis is carried at the segment level and its goal is to combine isolated letterlike segments into larger text-like structures, further referred to as elementary text clusters. At the second stage an iterative procedure is used to combine elementary clusters into larger text structures, called text chains. 5.1 Formation of elementary text clusters Three types of criteria, based on the horizontal alignment, relative height and color similarity, are used to create elementary clusters. The search starts from defining the left and right square neighborhoods of analyzed segments (such as segment A in Figure 2). The height and vertical location of such neighborhoods is identical with the height and vertical location of the analyzed segment. When the left or right neighbor was already found, only one neighborhood is examined. Each segment with a center of gravity located inside the left/right neighborhood is examined to check whether it is similar to the analyzed segment. Such selection of the candidate neighboring segments guarantees a reasonable tolerance to text distortions. First, the inter-segment relative height index γ is computed and compared with the upper and lower thresholds (determined in the experimental way) γ min γ = h a h c h a γ max where h a denotes the height of the analyzed segment and h c - the height of the compared segment. Second, the color similarity is examined using the weighted distance measure test δ = w 1 L a L c + w 2 ( āa ā c + b a b c ) δ max where L a, ā a, b a and L c, ā c, b c denote the mean Lab color coordinates of the analyzed segment and the compared segment, respectively EURASIP 274

4 From all segments that fulfill the similarity criteria described above, we choose the one that is closest to the analyzed segment - as a measure of closeness of two segments we use the distance between their centers of gravity. Only one segment in each neighborhood (if any) can be chosen. For each segment comprising the elementary text cluster the following coefficients, characterizing the slopes of the segments connecting the centre of gravity of the analyzed segment with the centres of gravity of its left/right neighbors, are evaluated (see Figure 2) m l = x x l, n l = y y l d l d l m r = x x r, n r = y y r d r d r d l = (x x l ) 2 + (y y l ) 2 d r = (x x r ) 2 + (y y r ) 2 Finally, for all segments that have both the left and right neighbors (as in Figure 3), the mean slope coefficients are computed m = m l + m r, n = n l + n r 2 2 Otherwise we set m = m l, n = n l (if only the left neighbor exists) or m = m r, n = n r (if only the right neighbor exists). 5.2 Formation of text chains At the second stage of clusterization, elementary clusters are grouped into text chains. Two elementary clusters can be considered for merging if their horizontal and vertical distances do not exceed certain limits. The horizontal and vertical distances between two clusters are defined as horizontal/vertical distances between centers of gravity of their closest, in the Euclidean sense, boundary (i.e. leftmost or rightmost) segments. It is required that the inter-cluster distance, measured in both directions, be smaller than half of the average height of each of the merged clusters. The average cluster height is calculated as an arithmetic mean of the heights of its component segments. In the case of three clusters, shown in Figure 3, in order to check whether the first cluster can be merged with the second cluster or the third one, one should check distances between centers of gravity of segments denoted as A, B and C. Since the vertical distance between segments A and C exceeds the threshold defined above, the third cluster must be rejected. On the contrary, since both horizontal and vertical distances between segments A and B are sufficiently small, the second cluster can be considered as a match for the first cluster. Every pair of clusters that passes the distance test, described above, is subject to a more detailed consistency check, based on the following weighted similarity measure D(k,l) = c 1 D Lab (k,l) + c 2 D h (k,l) + c 3 D mn (k,l) where k and l are the numbers of the compared clusters, D Lab (k,l) = L(k) L(l) + ā(k) ā(l) + b(k) b(l) Fig. 3. Formation of text chains. denotes the average (unweighted) distance in the Lab color space, based on comparison of the average Lab components determined for both clusters, D h (k,l) = h(k) h(l) denotes the absolute difference between the average cluster heights h(k) and h(l), and finally D mn (k,l) = m(k) m(l) + n(k) n(l) is the cluster slope similarity measure, based on comparison of the average m and n coefficients, determined for each cluster. 6. PRELIMINARY EXPERIMENTAL RESULTS AND FUTURE RESEARCH The proposed text detection algorithm was tested on a large number of photographs taken in the urban environment. Additionally, it was cross-checked using the benchmark database prepared for the ICDAR 2003 Robust Reading Competition. Since our system is still under development we focused on qualitative aspects of its performance. The two text-detection examples, shown in Figures 4 and 5, confirm that one of our main objectives - high tolerance to text slantness - was actually attained. To further increase system robustness and its detection efficiency we continue research in several directions, which include: resegmentation - to improve text readability, the detected text areas can be processed again using more sophisticated segmentation algorithms text rectification - OCRs have problem reading heavily slanted text; slant elimination can be performed easily using the segment and cluster slope parameters, described in Section 5.1. detection of large isolated text signs - the current version of the program rejects all isolated segments, even if they are letter/digit-shaped; while in the case of small segments this is usually a good strategy, for large segments it may result in loosing important pieces of information, such as single-digit bus and street numbers, for example EURASIP 275

5 Fig. 4. Original image and the results of text detection. REFERENCES [1] S. Ceranka and M. Niedźwiecki, Application of particle filtering in navigation system for the blind, 7th International Symposium on Signal Processing and its Applications, Paris, France, 2003, pp [2] S. Ceranka, GPS and dead reckoning based localization system for the blind, PhD Thesis (in Polish), Gdańsk University of Technology, [3] T. Strotthe et al., Mobility of blind and elderly people interacting with computers, National Institute for the Blind, report on the MOBIC project, [4] V. Gaudissart, S. Ferreira, C. Thillou and B. Gosselin, SYPOLE: a mobile assistant for the blind, Proc. 13th European Signal Processing Conference, EUSIPCO 2005, Antalya, Turkey, 2005, pp [5] S. Ferreira, C. Thilou and B. Gosselin, From picture to speech: an innovative application for embedded environment, Proc. of the 14th ProRISC Workshop on Circuits, Systems and Signal Processing, ProRISC 03, [6] N. Ezaki, M. Bulacu and L. Schomaker, Text detection from natural scene images: towards a system for visually impaired persons, Proc. 17th IEEE International Conference on Pattern Recognition, ICPR 04, 2004, pp Fig. 5. Original image and the results of text detection. [7] N. Ezaki, K. Kiyota, B.T. Minh, M. Bulacu and L. Schomaker, Improved text-detection methods for a camera-based text reading system for blind persons, Proc. 8th International Conference on Document Analysis and Recognition, ICDAR 05, 2005, pp [8] E.D. Haritaoglu and I. Haritaoglu, Real time image enhancement and segmentation for sign/text detection, Proc. IEEE International Conference on Image Processing, ICIP 03, 2003, pp. II993 II996. [9] J. Zhang, X. Chen, A. Hanneman, J. Yang and A. Waibel, A robust approach for recognition of text embedded in natural scenes,, Proc. IEEE International Conference on Pattern Recognition, ICPR 02, 2002, p [10] J. Gao and J. Yang, An adaptive algorithm for text detection from natural scenes, Proc. IEEE Computer Society CVPR 01, 2001, pp. II84 II89. [11] X. Chen and A.L. Yuille, Detecting and reading text in natural scenes, Proc IEEE Computer Society CVPR 04, 2004, pp. II366 II387. [12] X. Chen and A.L. Yuille, A time-efficient cascade for real-time object recognition with applications for the visually impaired, Proc IEEE Computer Society CVPR 05, 2005, p EURASIP 276

Text Detection in Indoor/Outdoor Scene Images

Text Detection in Indoor/Outdoor Scene Images Text Detection in Indoor/Outdoor Scene Images B. Gatos, I. Pratikakis, K. Kepene and S.J. Perantonis Computational Intelligence Laboratory, Institute of Informatics and Telecommunications, National Center

More information

Text Extraction from Natural Scene Images and Conversion to Audio in Smart Phone Applications

Text Extraction from Natural Scene Images and Conversion to Audio in Smart Phone Applications Text Extraction from Natural Scene Images and Conversion to Audio in Smart Phone Applications M. Prabaharan 1, K. Radha 2 M.E Student, Department of Computer Science and Engineering, Muthayammal Engineering

More information

Text-Tracking Wearable Camera System for the Blind

Text-Tracking Wearable Camera System for the Blind 2009 10th International Conference on Document Analysis and Recognition Text-Tracking Wearable Camera System for the Blind Hideaki Goto Cyberscience Center Tohoku University, Japan hgot @ isc.tohoku.ac.jp

More information

A Generalized Method to Solve Text-Based CAPTCHAs

A Generalized Method to Solve Text-Based CAPTCHAs A Generalized Method to Solve Text-Based CAPTCHAs Jason Ma, Bilal Badaoui, Emile Chamoun December 11, 2009 1 Abstract We present work in progress on the automated solving of text-based CAPTCHAs. Our method

More information

Scene Text Detection Using Machine Learning Classifiers

Scene Text Detection Using Machine Learning Classifiers 601 Scene Text Detection Using Machine Learning Classifiers Nafla C.N. 1, Sneha K. 2, Divya K.P. 3 1 (Department of CSE, RCET, Akkikkvu, Thrissur) 2 (Department of CSE, RCET, Akkikkvu, Thrissur) 3 (Department

More information

Figure 1: Workflow of object-based classification

Figure 1: Workflow of object-based classification Technical Specifications Object Analyst Object Analyst is an add-on package for Geomatica that provides tools for segmentation, classification, and feature extraction. Object Analyst includes an all-in-one

More information

An ICA based Approach for Complex Color Scene Text Binarization

An ICA based Approach for Complex Color Scene Text Binarization An ICA based Approach for Complex Color Scene Text Binarization Siddharth Kherada IIIT-Hyderabad, India siddharth.kherada@research.iiit.ac.in Anoop M. Namboodiri IIIT-Hyderabad, India anoop@iiit.ac.in

More information

Image Segmentation for Image Object Extraction

Image Segmentation for Image Object Extraction Image Segmentation for Image Object Extraction Rohit Kamble, Keshav Kaul # Computer Department, Vishwakarma Institute of Information Technology, Pune kamble.rohit@hotmail.com, kaul.keshav@gmail.com ABSTRACT

More information

Texture Segmentation by Windowed Projection

Texture Segmentation by Windowed Projection Texture Segmentation by Windowed Projection 1, 2 Fan-Chen Tseng, 2 Ching-Chi Hsu, 2 Chiou-Shann Fuh 1 Department of Electronic Engineering National I-Lan Institute of Technology e-mail : fctseng@ccmail.ilantech.edu.tw

More information

2D rendering takes a photo of the 2D scene with a virtual camera that selects an axis aligned rectangle from the scene. The photograph is placed into

2D rendering takes a photo of the 2D scene with a virtual camera that selects an axis aligned rectangle from the scene. The photograph is placed into 2D rendering takes a photo of the 2D scene with a virtual camera that selects an axis aligned rectangle from the scene. The photograph is placed into the viewport of the current application window. A pixel

More information

Mobile Camera Based Text Detection and Translation

Mobile Camera Based Text Detection and Translation Mobile Camera Based Text Detection and Translation Derek Ma Qiuhau Lin Tong Zhang Department of Electrical EngineeringDepartment of Electrical EngineeringDepartment of Mechanical Engineering Email: derekxm@stanford.edu

More information

Text Information Extraction And Analysis From Images Using Digital Image Processing Techniques

Text Information Extraction And Analysis From Images Using Digital Image Processing Techniques Text Information Extraction And Analysis From Images Using Digital Image Processing Techniques Partha Sarathi Giri Department of Electronics and Communication, M.E.M.S, Balasore, Odisha Abstract Text data

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

Conspicuous Character Patterns

Conspicuous Character Patterns Conspicuous Character Patterns Seiichi Uchida Kyushu Univ., Japan Ryoji Hattori Masakazu Iwamura Kyushu Univ., Japan Osaka Pref. Univ., Japan Koichi Kise Osaka Pref. Univ., Japan Shinichiro Omachi Tohoku

More information

Text Area Detection from Video Frames

Text Area Detection from Video Frames Text Area Detection from Video Frames 1 Text Area Detection from Video Frames Xiangrong Chen, Hongjiang Zhang Microsoft Research China chxr@yahoo.com, hjzhang@microsoft.com Abstract. Text area detection

More information

I. INTRODUCTION. Figure-1 Basic block of text analysis

I. INTRODUCTION. Figure-1 Basic block of text analysis ISSN: 2349-7637 (Online) (RHIMRJ) Research Paper Available online at: www.rhimrj.com Detection and Localization of Texts from Natural Scene Images: A Hybrid Approach Priyanka Muchhadiya Post Graduate Fellow,

More information

Chapter 3 Image Registration. Chapter 3 Image Registration

Chapter 3 Image Registration. Chapter 3 Image Registration Chapter 3 Image Registration Distributed Algorithms for Introduction (1) Definition: Image Registration Input: 2 images of the same scene but taken from different perspectives Goal: Identify transformation

More information

Scanner Parameter Estimation Using Bilevel Scans of Star Charts

Scanner Parameter Estimation Using Bilevel Scans of Star Charts ICDAR, Seattle WA September Scanner Parameter Estimation Using Bilevel Scans of Star Charts Elisa H. Barney Smith Electrical and Computer Engineering Department Boise State University, Boise, Idaho 8375

More information

COMPARATIVE STUDY OF DIFFERENT APPROACHES FOR EFFICIENT RECTIFICATION UNDER GENERAL MOTION

COMPARATIVE STUDY OF DIFFERENT APPROACHES FOR EFFICIENT RECTIFICATION UNDER GENERAL MOTION COMPARATIVE STUDY OF DIFFERENT APPROACHES FOR EFFICIENT RECTIFICATION UNDER GENERAL MOTION Mr.V.SRINIVASA RAO 1 Prof.A.SATYA KALYAN 2 DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING PRASAD V POTLURI SIDDHARTHA

More information

Automatic Video Caption Detection and Extraction in the DCT Compressed Domain

Automatic Video Caption Detection and Extraction in the DCT Compressed Domain Automatic Video Caption Detection and Extraction in the DCT Compressed Domain Chin-Fu Tsao 1, Yu-Hao Chen 1, Jin-Hau Kuo 1, Chia-wei Lin 1, and Ja-Ling Wu 1,2 1 Communication and Multimedia Laboratory,

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

Bus Detection and recognition for visually impaired people

Bus Detection and recognition for visually impaired people Bus Detection and recognition for visually impaired people Hangrong Pan, Chucai Yi, and Yingli Tian The City College of New York The Graduate Center The City University of New York MAP4VIP Outline Motivation

More information

2 Proposed Methodology

2 Proposed Methodology 3rd International Conference on Multimedia Technology(ICMT 2013) Object Detection in Image with Complex Background Dong Li, Yali Li, Fei He, Shengjin Wang 1 State Key Laboratory of Intelligent Technology

More information

Cs : Computer Vision Final Project Report

Cs : Computer Vision Final Project Report Cs 600.461: Computer Vision Final Project Report Giancarlo Troni gtroni@jhu.edu Raphael Sznitman sznitman@jhu.edu Abstract Given a Youtube video of a busy street intersection, our task is to detect, track,

More information

Unique Journal of Engineering and Advanced Sciences Available online: Research Article

Unique Journal of Engineering and Advanced Sciences Available online:  Research Article ISSN 2348-375X Unique Journal of Engineering and Advanced Sciences Available online: www.ujconline.net Research Article DETECTION AND RECOGNITION OF THE TEXT THROUGH CONNECTED COMPONENT CLUSTERING AND

More information

Consistent Line Clusters for Building Recognition in CBIR

Consistent Line Clusters for Building Recognition in CBIR Consistent Line Clusters for Building Recognition in CBIR Yi Li and Linda G. Shapiro Department of Computer Science and Engineering University of Washington Seattle, WA 98195-250 shapiro,yi @cs.washington.edu

More information

arxiv: v1 [cs.cv] 28 Sep 2018

arxiv: v1 [cs.cv] 28 Sep 2018 Camera Pose Estimation from Sequence of Calibrated Images arxiv:1809.11066v1 [cs.cv] 28 Sep 2018 Jacek Komorowski 1 and Przemyslaw Rokita 2 1 Maria Curie-Sklodowska University, Institute of Computer Science,

More information

Segmentation Framework for Multi-Oriented Text Detection and Recognition

Segmentation Framework for Multi-Oriented Text Detection and Recognition Segmentation Framework for Multi-Oriented Text Detection and Recognition Shashi Kant, Sini Shibu Department of Computer Science and Engineering, NRI-IIST, Bhopal Abstract - Here in this paper a new and

More information

Efficient Acquisition of Human Existence Priors from Motion Trajectories

Efficient Acquisition of Human Existence Priors from Motion Trajectories Efficient Acquisition of Human Existence Priors from Motion Trajectories Hitoshi Habe Hidehito Nakagawa Masatsugu Kidode Graduate School of Information Science, Nara Institute of Science and Technology

More information

A Laplacian Based Novel Approach to Efficient Text Localization in Grayscale Images

A Laplacian Based Novel Approach to Efficient Text Localization in Grayscale Images A Laplacian Based Novel Approach to Efficient Text Localization in Grayscale Images Karthik Ram K.V & Mahantesh K Department of Electronics and Communication Engineering, SJB Institute of Technology, Bangalore,

More information

A GENETIC ALGORITHM FOR MOTION DETECTION

A GENETIC ALGORITHM FOR MOTION DETECTION A GENETIC ALGORITHM FOR MOTION DETECTION Jarosław Mamica, Tomasz Walkowiak Institute of Engineering Cybernetics, Wrocław University of Technology ul. Janiszewskiego 11/17, 50-37 Wrocław, POLAND, Phone:

More information

A Noise-Robust and Adaptive Image Segmentation Method based on Splitting and Merging method

A Noise-Robust and Adaptive Image Segmentation Method based on Splitting and Merging method A Noise-Robust and Adaptive Image Segmentation Method based on Splitting and Merging method Ryu Hyunki, Lee HaengSuk Kyungpook Research Institute of Vehicle Embedded Tech. 97-70, Myeongsan-gil, YeongCheon,

More information

Introduction to Medical Imaging (5XSA0) Module 5

Introduction to Medical Imaging (5XSA0) Module 5 Introduction to Medical Imaging (5XSA0) Module 5 Segmentation Jungong Han, Dirk Farin, Sveta Zinger ( s.zinger@tue.nl ) 1 Outline Introduction Color Segmentation region-growing region-merging watershed

More information

DESIGNING A REAL TIME SYSTEM FOR CAR NUMBER DETECTION USING DISCRETE HOPFIELD NETWORK

DESIGNING A REAL TIME SYSTEM FOR CAR NUMBER DETECTION USING DISCRETE HOPFIELD NETWORK DESIGNING A REAL TIME SYSTEM FOR CAR NUMBER DETECTION USING DISCRETE HOPFIELD NETWORK A.BANERJEE 1, K.BASU 2 and A.KONAR 3 COMPUTER VISION AND ROBOTICS LAB ELECTRONICS AND TELECOMMUNICATION ENGG JADAVPUR

More information

Research on QR Code Image Pre-processing Algorithm under Complex Background

Research on QR Code Image Pre-processing Algorithm under Complex Background Scientific Journal of Information Engineering May 207, Volume 7, Issue, PP.-7 Research on QR Code Image Pre-processing Algorithm under Complex Background Lei Liu, Lin-li Zhou, Huifang Bao. Institute of

More information

Binarization of Color Character Strings in Scene Images Using K-means Clustering and Support Vector Machines

Binarization of Color Character Strings in Scene Images Using K-means Clustering and Support Vector Machines 2011 International Conference on Document Analysis and Recognition Binarization of Color Character Strings in Scene Images Using K-means Clustering and Support Vector Machines Toru Wakahara Kohei Kita

More information

An Approach to Detect Text and Caption in Video

An Approach to Detect Text and Caption in Video An Approach to Detect Text and Caption in Video Miss Megha Khokhra 1 M.E Student Electronics and Communication Department, Kalol Institute of Technology, Gujarat, India ABSTRACT The video image spitted

More information

EE795: Computer Vision and Intelligent Systems

EE795: Computer Vision and Intelligent Systems EE795: Computer Vision and Intelligent Systems Spring 2012 TTh 17:30-18:45 FDH 204 Lecture 14 130307 http://www.ee.unlv.edu/~b1morris/ecg795/ 2 Outline Review Stereo Dense Motion Estimation Translational

More information

TEXTURE BASED TEXT DETECTION IN NATURAL SCENE IMAGES: A HELP TO BLIND AND VISUALLY IMPAIRED PERSONS

TEXTURE BASED TEXT DETECTION IN NATURAL SCENE IMAGES: A HELP TO BLIND AND VISUALLY IMPAIRED PERSONS Conference & Workshop on Assistive Technologies for People with Vision & Hearing Impairments Assistive Technology for All Ages CVHI 2007, M.A. Hersh (ed.) TEXTURE BASED TEXT DETECTION IN NATURAL SCENE

More information

Object-Based Classification & ecognition. Zutao Ouyang 11/17/2015

Object-Based Classification & ecognition. Zutao Ouyang 11/17/2015 Object-Based Classification & ecognition Zutao Ouyang 11/17/2015 What is Object-Based Classification The object based image analysis approach delineates segments of homogeneous image areas (i.e., objects)

More information

Compression of RADARSAT Data with Block Adaptive Wavelets Abstract: 1. Introduction

Compression of RADARSAT Data with Block Adaptive Wavelets Abstract: 1. Introduction Compression of RADARSAT Data with Block Adaptive Wavelets Ian Cumming and Jing Wang Department of Electrical and Computer Engineering The University of British Columbia 2356 Main Mall, Vancouver, BC, Canada

More information

ABSTRACT 1. INTRODUCTION 2. RELATED WORK

ABSTRACT 1. INTRODUCTION 2. RELATED WORK Improving text recognition by distinguishing scene and overlay text Bernhard Quehl, Haojin Yang, Harald Sack Hasso Plattner Institute, Potsdam, Germany Email: {bernhard.quehl, haojin.yang, harald.sack}@hpi.de

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

Indian Multi-Script Full Pin-code String Recognition for Postal Automation

Indian Multi-Script Full Pin-code String Recognition for Postal Automation 2009 10th International Conference on Document Analysis and Recognition Indian Multi-Script Full Pin-code String Recognition for Postal Automation U. Pal 1, R. K. Roy 1, K. Roy 2 and F. Kimura 3 1 Computer

More information

Solving Word Jumbles

Solving Word Jumbles Solving Word Jumbles Debabrata Sengupta, Abhishek Sharma Department of Electrical Engineering, Stanford University { dsgupta, abhisheksharma }@stanford.edu Abstract In this report we propose an algorithm

More information

Histogram and watershed based segmentation of color images

Histogram and watershed based segmentation of color images Histogram and watershed based segmentation of color images O. Lezoray H. Cardot LUSAC EA 2607 IUT Saint-Lô, 120 rue de l'exode, 50000 Saint-Lô, FRANCE Abstract A novel method for color image segmentation

More information

Use of Shape Deformation to Seamlessly Stitch Historical Document Images

Use of Shape Deformation to Seamlessly Stitch Historical Document Images Use of Shape Deformation to Seamlessly Stitch Historical Document Images Wei Liu Wei Fan Li Chen Jun Sun Satoshi Naoi In China, efforts are being made to preserve historical documents in the form of digital

More information

Text Block Detection and Segmentation for Mobile Robot Vision System Applications

Text Block Detection and Segmentation for Mobile Robot Vision System Applications Proc. of Int. Conf. onmultimedia Processing, Communication and Info. Tech., MPCIT Text Block Detection and Segmentation for Mobile Robot Vision System Applications Too Boaz Kipyego and Prabhakar C. J.

More information

Recognition of Gurmukhi Text from Sign Board Images Captured from Mobile Camera

Recognition of Gurmukhi Text from Sign Board Images Captured from Mobile Camera International Journal of Information & Computation Technology. ISSN 0974-2239 Volume 4, Number 17 (2014), pp. 1839-1845 International Research Publications House http://www. irphouse.com Recognition of

More information

Topic 4 Image Segmentation

Topic 4 Image Segmentation Topic 4 Image Segmentation What is Segmentation? Why? Segmentation important contributing factor to the success of an automated image analysis process What is Image Analysis: Processing images to derive

More information

CORRELATION BASED CAR NUMBER PLATE EXTRACTION SYSTEM

CORRELATION BASED CAR NUMBER PLATE EXTRACTION SYSTEM CORRELATION BASED CAR NUMBER PLATE EXTRACTION SYSTEM 1 PHYO THET KHIN, 2 LAI LAI WIN KYI 1,2 Department of Information Technology, Mandalay Technological University The Republic of the Union of Myanmar

More information

Previously. Part-based and local feature models for generic object recognition. Bag-of-words model 4/20/2011

Previously. Part-based and local feature models for generic object recognition. Bag-of-words model 4/20/2011 Previously Part-based and local feature models for generic object recognition Wed, April 20 UT-Austin Discriminative classifiers Boosting Nearest neighbors Support vector machines Useful for object recognition

More information

Detecting and Reading Text in Natural Scenes

Detecting and Reading Text in Natural Scenes October 19, 2004 X. Chen, A. L. Yuille Outline Outline Goals Example Main Ideas Results Goals Outline Goals Example Main Ideas Results Given an image of an outdoor scene, goals are to: Identify regions

More information

Detection of a Single Hand Shape in the Foreground of Still Images

Detection of a Single Hand Shape in the Foreground of Still Images CS229 Project Final Report Detection of a Single Hand Shape in the Foreground of Still Images Toan Tran (dtoan@stanford.edu) 1. Introduction This paper is about an image detection system that can detect

More information

Robotics Programming Laboratory

Robotics Programming Laboratory Chair of Software Engineering Robotics Programming Laboratory Bertrand Meyer Jiwon Shin Lecture 8: Robot Perception Perception http://pascallin.ecs.soton.ac.uk/challenges/voc/databases.html#caltech car

More information

[10] Industrial DataMatrix barcodes recognition with a random tilt and rotating the camera

[10] Industrial DataMatrix barcodes recognition with a random tilt and rotating the camera [10] Industrial DataMatrix barcodes recognition with a random tilt and rotating the camera Image processing, pattern recognition 865 Kruchinin A.Yu. Orenburg State University IntBuSoft Ltd Abstract The

More information

OTCYMIST: Otsu-Canny Minimal Spanning Tree for Born-Digital Images

OTCYMIST: Otsu-Canny Minimal Spanning Tree for Born-Digital Images OTCYMIST: Otsu-Canny Minimal Spanning Tree for Born-Digital Images Deepak Kumar and A G Ramakrishnan Medical Intelligence and Language Engineering Laboratory Department of Electrical Engineering, Indian

More information

Sketchable Histograms of Oriented Gradients for Object Detection

Sketchable Histograms of Oriented Gradients for Object Detection Sketchable Histograms of Oriented Gradients for Object Detection No Author Given No Institute Given Abstract. In this paper we investigate a new representation approach for visual object recognition. The

More information

TEVI: Text Extraction for Video Indexing

TEVI: Text Extraction for Video Indexing TEVI: Text Extraction for Video Indexing Hichem KARRAY, Mohamed SALAH, Adel M. ALIMI REGIM: Research Group on Intelligent Machines, EIS, University of Sfax, Tunisia hichem.karray@ieee.org mohamed_salah@laposte.net

More information

A System for Joining and Recognition of Broken Bangla Numerals for Indian Postal Automation

A System for Joining and Recognition of Broken Bangla Numerals for Indian Postal Automation A System for Joining and Recognition of Broken Bangla Numerals for Indian Postal Automation K. Roy, U. Pal and B. B. Chaudhuri CVPR Unit; Indian Statistical Institute, Kolkata-108; India umapada@isical.ac.in

More information

Scene Text Recognition using Co-occurrence of Histogram of Oriented Gradients

Scene Text Recognition using Co-occurrence of Histogram of Oriented Gradients 203 2th International Conference on Document Analysis and Recognition Scene Text Recognition using Co-occurrence of Histogram of Oriented Gradients Shangxuan Tian, Shijian Lu, Bolan Su and Chew Lim Tan

More information

TEXT EXTRACTION FROM AN IMAGE BY USING DIGITAL IMAGE PROCESSING

TEXT EXTRACTION FROM AN IMAGE BY USING DIGITAL IMAGE PROCESSING TEXT EXTRACTION FROM AN IMAGE BY USING DIGITAL IMAGE PROCESSING Praveen Choudhary Assistant Professor, S.S. Jain Subodh P.G. College, Jaipur, India Prateeshvip79@gmail.com Dr. Vipin Kumar Jain Assistant

More information

SYDE Winter 2011 Introduction to Pattern Recognition. Clustering

SYDE Winter 2011 Introduction to Pattern Recognition. Clustering SYDE 372 - Winter 2011 Introduction to Pattern Recognition Clustering Alexander Wong Department of Systems Design Engineering University of Waterloo Outline 1 2 3 4 5 All the approaches we have learned

More information

International Journal of Advance Research in Engineering, Science & Technology

International Journal of Advance Research in Engineering, Science & Technology Impact Factor (SJIF): 4.542 International Journal of Advance Research in Engineering, Science & Technology e-issn: 2393-9877, p-issn: 2394-2444 Volume 4, Issue 4, April-2017 A Simple Effective Algorithm

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

OPTIMIZING A VIDEO PREPROCESSOR FOR OCR. MR IBM Systems Dev Rochester, elopment Division Minnesota

OPTIMIZING A VIDEO PREPROCESSOR FOR OCR. MR IBM Systems Dev Rochester, elopment Division Minnesota OPTIMIZING A VIDEO PREPROCESSOR FOR OCR MR IBM Systems Dev Rochester, elopment Division Minnesota Summary This paper describes how optimal video preprocessor performance can be achieved using a software

More information

Vision. OCR and OCV Application Guide OCR and OCV Application Guide 1/14

Vision. OCR and OCV Application Guide OCR and OCV Application Guide 1/14 Vision OCR and OCV Application Guide 1.00 OCR and OCV Application Guide 1/14 General considerations on OCR Encoded information into text and codes can be automatically extracted through a 2D imager device.

More information

Research on Recognition and Classification of Moving Objects in Mixed Traffic Based on Video Detection

Research on Recognition and Classification of Moving Objects in Mixed Traffic Based on Video Detection Hu, Qu, Li and Wang 1 Research on Recognition and Classification of Moving Objects in Mixed Traffic Based on Video Detection Hongyu Hu (corresponding author) College of Transportation, Jilin University,

More information

Report: Reducing the error rate of a Cat classifier

Report: Reducing the error rate of a Cat classifier Report: Reducing the error rate of a Cat classifier Raphael Sznitman 6 August, 2007 Abstract The following report discusses my work at the IDIAP from 06.2007 to 08.2007. This work had for objective to

More information

Enhanced Hemisphere Concept for Color Pixel Classification

Enhanced Hemisphere Concept for Color Pixel Classification 2016 International Conference on Multimedia Systems and Signal Processing Enhanced Hemisphere Concept for Color Pixel Classification Van Ng Graduate School of Information Sciences Tohoku University Sendai,

More information

TEXT SEGMENTATION ON PHOTOREALISTIC IMAGES

TEXT SEGMENTATION ON PHOTOREALISTIC IMAGES TEXT SEGMENTATION ON PHOTOREALISTIC IMAGES Valery Grishkin a, Alexander Ebral b, Nikolai Stepenko c, Jean Sene d Saint Petersburg State University, 7 9 Universitetskaya nab., Saint Petersburg, 199034,

More information

Vehicle Dimensions Estimation Scheme Using AAM on Stereoscopic Video

Vehicle Dimensions Estimation Scheme Using AAM on Stereoscopic Video Workshop on Vehicle Retrieval in Surveillance (VRS) in conjunction with 2013 10th IEEE International Conference on Advanced Video and Signal Based Surveillance Vehicle Dimensions Estimation Scheme Using

More information

Colour Segmentation-based Computation of Dense Optical Flow with Application to Video Object Segmentation

Colour Segmentation-based Computation of Dense Optical Flow with Application to Video Object Segmentation ÖGAI Journal 24/1 11 Colour Segmentation-based Computation of Dense Optical Flow with Application to Video Object Segmentation Michael Bleyer, Margrit Gelautz, Christoph Rhemann Vienna University of Technology

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

Texture Sensitive Image Inpainting after Object Morphing

Texture Sensitive Image Inpainting after Object Morphing Texture Sensitive Image Inpainting after Object Morphing Yin Chieh Liu and Yi-Leh Wu Department of Computer Science and Information Engineering National Taiwan University of Science and Technology, Taiwan

More information

Optical Microscopy and Metallography: Recognition of Metal Structures using Customizable Automatic Qualifiers

Optical Microscopy and Metallography: Recognition of Metal Structures using Customizable Automatic Qualifiers ECNDT 2006 - Poster 68 Optical Microscopy and Metallography: Recognition of Metal Structures using Customizable Automatic Qualifiers Mikhail V. FILINOV, Andrey S. FURSOV, Association SPEKTR_GROUP, Moscow,

More information

A Local Statistics Based Region Growing Segmentation Method for Ultrasound Medical Images

A Local Statistics Based Region Growing Segmentation Method for Ultrasound Medical Images A Local Statistics Based Region Growing Segmentation Method for Ultrasound Medical Images Ashish Thakur Radhey Shyam Anand * Abstract This paper presents the region based segmentation method for ultrasound

More information

Segmentation of Images

Segmentation of Images Segmentation of Images SEGMENTATION If an image has been preprocessed appropriately to remove noise and artifacts, segmentation is often the key step in interpreting the image. Image segmentation is a

More information

Image retrieval based on region shape similarity

Image retrieval based on region shape similarity Image retrieval based on region shape similarity Cheng Chang Liu Wenyin Hongjiang Zhang Microsoft Research China, 49 Zhichun Road, Beijing 8, China {wyliu, hjzhang}@microsoft.com ABSTRACT This paper presents

More information

EE 701 ROBOT VISION. Segmentation

EE 701 ROBOT VISION. Segmentation EE 701 ROBOT VISION Regions and Image Segmentation Histogram-based Segmentation Automatic Thresholding K-means Clustering Spatial Coherence Merging and Splitting Graph Theoretic Segmentation Region Growing

More information

SCENE TEXT BINARIZATION AND RECOGNITION

SCENE TEXT BINARIZATION AND RECOGNITION Chapter 5 SCENE TEXT BINARIZATION AND RECOGNITION 5.1 BACKGROUND In the previous chapter, detection of text lines from scene images using run length based method and also elimination of false positives

More information

Region-based Segmentation and Object Detection

Region-based Segmentation and Object Detection Region-based Segmentation and Object Detection Stephen Gould Tianshi Gao Daphne Koller Presented at NIPS 2009 Discussion and Slides by Eric Wang April 23, 2010 Outline Introduction Model Overview Model

More information

Content-based Image and Video Retrieval. Image Segmentation

Content-based Image and Video Retrieval. Image Segmentation Content-based Image and Video Retrieval Vorlesung, SS 2011 Image Segmentation 2.5.2011 / 9.5.2011 Image Segmentation One of the key problem in computer vision Identification of homogenous region in the

More information

Research on-board LIDAR point cloud data pretreatment

Research on-board LIDAR point cloud data pretreatment Acta Technica 62, No. 3B/2017, 1 16 c 2017 Institute of Thermomechanics CAS, v.v.i. Research on-board LIDAR point cloud data pretreatment Peng Cang 1, Zhenglin Yu 1, Bo Yu 2, 3 Abstract. In view of the

More information

A Novel Approach to Image Segmentation for Traffic Sign Recognition Jon Jay Hack and Sidd Jagadish

A Novel Approach to Image Segmentation for Traffic Sign Recognition Jon Jay Hack and Sidd Jagadish A Novel Approach to Image Segmentation for Traffic Sign Recognition Jon Jay Hack and Sidd Jagadish Introduction/Motivation: As autonomous vehicles, such as Google s self-driving car, have recently become

More information

Textural Features for Image Database Retrieval

Textural Features for Image Database Retrieval Textural Features for Image Database Retrieval Selim Aksoy and Robert M. Haralick Intelligent Systems Laboratory Department of Electrical Engineering University of Washington Seattle, WA 98195-2500 {aksoy,haralick}@@isl.ee.washington.edu

More information

Connected Component Analysis and Change Detection for Images

Connected Component Analysis and Change Detection for Images Connected Component Analysis and Change Detection for Images Prasad S.Halgaonkar Department of Computer Engg, MITCOE Pune University, India Abstract Detection of the region of change in images of a particular

More information

Mobile Human Detection Systems based on Sliding Windows Approach-A Review

Mobile Human Detection Systems based on Sliding Windows Approach-A Review Mobile Human Detection Systems based on Sliding Windows Approach-A Review Seminar: Mobile Human detection systems Njieutcheu Tassi cedrique Rovile Department of Computer Engineering University of Heidelberg

More information

Network Snakes for the Segmentation of Adjacent Cells in Confocal Images

Network Snakes for the Segmentation of Adjacent Cells in Confocal Images Network Snakes for the Segmentation of Adjacent Cells in Confocal Images Matthias Butenuth 1 and Fritz Jetzek 2 1 Institut für Photogrammetrie und GeoInformation, Leibniz Universität Hannover, 30167 Hannover

More information

Mobile Visual Search with Word-HOG Descriptors

Mobile Visual Search with Word-HOG Descriptors Mobile Visual Search with Word-HOG Descriptors Sam S. Tsai, Huizhong Chen, David M. Chen, and Bernd Girod Department of Electrical Engineering, Stanford University, Stanford, CA, 9435 sstsai@alumni.stanford.edu,

More information

Isolated Handwritten Words Segmentation Techniques in Gurmukhi Script

Isolated Handwritten Words Segmentation Techniques in Gurmukhi Script Isolated Handwritten Words Segmentation Techniques in Gurmukhi Script Galaxy Bansal Dharamveer Sharma ABSTRACT Segmentation of handwritten words is a challenging task primarily because of structural features

More information

A Fast and Accurate Eyelids and Eyelashes Detection Approach for Iris Segmentation

A Fast and Accurate Eyelids and Eyelashes Detection Approach for Iris Segmentation A Fast and Accurate Eyelids and Eyelashes Detection Approach for Iris Segmentation Walid Aydi, Lotfi Kamoun, Nouri Masmoudi Department of Electrical National Engineering School of Sfax Sfax University

More information

TCG-Based Multi-Bend Bus Driven Floorplanning

TCG-Based Multi-Bend Bus Driven Floorplanning TCG-Based Multi-Bend Bus Driven Floorplanning Tilen Ma Department of CSE The Chinese University of Hong Kong Shatin, N.T. Hong Kong Evangeline F.Y. Young Department of CSE The Chinese University of Hong

More information

COMPARISON OF SOME EFFICIENT METHODS OF CONTOUR COMPRESSION

COMPARISON OF SOME EFFICIENT METHODS OF CONTOUR COMPRESSION COMPARISON OF SOME EFFICIENT METHODS OF CONTOUR COMPRESSION REMIGIUSZ BARAN ANDRZEJ DZIECH 2 Department of Electronics and Intelligent Systems, Kielce University of Technology, POLAND 2 AGH University

More information

A Labeling Approach for Mixed Document Blocks. A. Bela d and O. T. Akindele. Crin-Cnrs/Inria-Lorraine, B timent LORIA, Campus Scientique, B.P.

A Labeling Approach for Mixed Document Blocks. A. Bela d and O. T. Akindele. Crin-Cnrs/Inria-Lorraine, B timent LORIA, Campus Scientique, B.P. A Labeling Approach for Mixed Document Blocks A. Bela d and O. T. Akindele Crin-Cnrs/Inria-Lorraine, B timent LORIA, Campus Scientique, B.P. 39, 54506 Vand uvre-l s-nancy Cedex. France. Abstract A block

More information

User s Manual. A Revolution in OCR Software

User s Manual. A Revolution in OCR Software User s Manual A Revolution in OCR Software Microsoft, Windows, Windows NT, Windows 2000, Windows XP, Visual Basic, Microsoft.NET, Visual C++, Visual C#, and ActiveX are either trademarks or registered

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

Recognition of Multiple Characters in a Scene Image Using Arrangement of Local Features

Recognition of Multiple Characters in a Scene Image Using Arrangement of Local Features 2011 International Conference on Document Analysis and Recognition Recognition of Multiple Characters in a Scene Image Using Arrangement of Local Features Masakazu Iwamura, Takuya Kobayashi, and Koichi

More information

A Combined Method for On-Line Signature Verification

A Combined Method for On-Line Signature Verification BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 14, No 2 Sofia 2014 Print ISSN: 1311-9702; Online ISSN: 1314-4081 DOI: 10.2478/cait-2014-0022 A Combined Method for On-Line

More information

Moving Object Segmentation Method Based on Motion Information Classification by X-means and Spatial Region Segmentation

Moving Object Segmentation Method Based on Motion Information Classification by X-means and Spatial Region Segmentation IJCSNS International Journal of Computer Science and Network Security, VOL.13 No.11, November 2013 1 Moving Object Segmentation Method Based on Motion Information Classification by X-means and Spatial

More information