Evaluation of several binarization techniques for old Arabic documents images

Size: px
Start display at page:

Download "Evaluation of several binarization techniques for old Arabic documents images"

Transcription

1 Evaluation of several binarization techniques for old Arabic documents images Abderrahmane KEFALI (1), Toufik SARI (2), Mokhtar SELLAMI (1) (1) Laboratoire de Recherche en Informatique (LRI), Département Informatique, BP 12 - Université Badji Mokhtar - Annaba, 23000, Algérie (2) Laboratoire de Gestion Electronique de Documents (LabGed), Département Informatique, BP 12 - Université Badji Mokhtar - Annaba, 23000, Algérie {abderahmane.kefali, sellami}@lri-annaba.net sari@labged.net Abstract Binarization is an important stage in all process of images processing and analysis. Currently, lot of techniques were proposed in the literature for binarizing colored and gray-level images, of which each one is appropriate to a particular type of images, but unfortunately no one among them surpassed in the case of the binarisation of old manuscript documents. These lasts are characterized by their bad quality due to the various deteriorations suffered in the life cycle of documents and to materials used for their storage. This paper presents an evaluation of binarization techniques frequently cited in the literature. In the first stage we studied, implemented and tested twelve binarisation algorithms over old Arabic manuscripts document. We however tried to compare them by using a new method based on features extracted from each image. The goal of our work is in a middle term to do a comparative study and can choose thereafter the best binarisation algorithm to use it in our Arabic documents retrieval system. Key words: thresholding, binarization, old Arabic documents. 1. Introduction Binarization is often the first stage in all systems of images (and more particularly documents images) processing and analysis [1][2]. We can cite here: Optical character Recognition (OCR), word spotting and indexation. It has as goal to decrease the quantity of information present in the image and to guard only relevant data, which allows as to use simple analysis methods face to face of gray-level or colored images [3]. Performances of the following steps in the documents analysis systems depends highly on the results of the used binarization algorithm, it is wherefore the binarization step must be performed the most accurately possible. It must on the one hand; conserve the maximum of information and details present in the entered image, and in the other hand it must eliminate the noise superposed to the original data. In fact, the problem of documents binarization is older than documents images analysis itself. For that, a very important number of binarization techniques were proposed in the literature which each one between them is appropriate to a particular type of images. Indeed, a lot of techniques were proposed in the two lasts decades for binarizing gray-level documents images, others are applicable directly over colored images [4], but unfortunately, no one between them showed powerful for the binarization of degraded old documents, characterized by a complex structure, illumination variety, presence of stains and holes, transparence effect etc. We find in the literature several works interested by the comparison of binarization methods over several types of images, citing for example [1] [7] [21] [22] [25]. In this paper, we tackle with the binarization of old Arabic documents images. For that, we implement twelve binarization algorithms, the most frequently cited in the literature, and we apply them over a set of degraded old Arabic documents images in order to evaluate their performances. For the evaluation, we use a new method consists to compare features extracted from images instead of the comparison between these images. We interest in this study by the binarization of gray-level documents only, because, the most of colored documents may be converted easily and faithfully to gray-level. The rest of this paper is organized as follow: in section 2, we cite the families of binarization methods and we give a state of the art of the studied binarization methods. Afterwards in the section 3, we present our test collection composed of old Arabic documents images, the obtained results, and the evaluation of these results. Finally, the conclusion will be in the section State of the art

2 According to several research works [5][6], the binarization techniques gray-level images can be classified in two categories: global thresholding where one threshold is used in the whole image to separate it in two classes (text and background), and local thresholding where the threshold values are determined locally, pixel by pixel or region by region. Other [3], added a third group of hybrid methods. These methods combine global and local information for assign the pixels to one of two classes (text or background). In [7], the binarisation methods were classified according to the information that they exploit in six categories: methods based on histogram shape, clustering, entropy, object attributes, spatial methods and finally local methods. In this section, we consider the first classification (global and local methods) and we present a little state of the art of several binarization methods the most frequently cited in the literature. For global thresholding methods, we choose: fixed global thresholding [8], Otsu method [9], ISODATA [10], Kapur [11], Cheng and Chen [12], Li-Lee [13], with a global version of the Nick method [6]. And for local methods, we choose the Bernsen method [14], Niblack [15], Sauvola [16], Wolf [17], Nick [6] and the Fuzzy hierarchical segmentation of Gadi and Benslimane [18] Global methods Fixed global thresholding It is the most simple binarization technique [8]; it consists to compare the gray-level of each pixel x i of the image with a global threshold T (for example 127). We note b i the new value of the pixel, thresholding is given by the following expression: b i = 255 if x i T and b i = 0 if x i < T Otsu method Otsu method [9] is one of the famous global methods. This method tries to find the threshold T which separates the gray-level histogram in an optimal way into two segments (which maximize the inter-segments variance or which minimize the intra-segments variance). We note here that the calculation of inter-classes or intra-classes variances is based on the normalized histogram of the image H= [h 0...h 255 ] where h i =1. The inter-classes variance for each pixel t is given by: σ inter 2 = q 1 (t) q 2 (t) [µ 1 (t)-µ 2 (t)] 2 Where: μ 1 t = 1 q 1 t q 1 t = t 1 i=0 h i i and μ 2 t = 1 q 2 t t 1 i=0 h(i) and q 2 t = 255 i=t 255 i=t h(i) h i i The limit of this method is that it applicable only when the image is bimodal (the histogram contain two peaks). Wolf and Jolion [19] proposed to use a two-thresholds version of the Otsu method for binarizing video images, where the first threshold k haut correspond to the threshold calculated by the Otsu method, and the second threshold k bas is calculated from the threshold k haut and the first mode of the histogram: k bas = m (k haut - m 0 ). Thus deciding the membership of pixels to one of Object or background classes is modified as follow: I x,y < k bas B x,y = 0 I x,y > k haut B x,y = 255 k bas > I x,y > k haut B x,y = 255 if exist a way to a pixel I u,v > k haut 0 else

3 Such as I x,y is the gray-level of the pixel (x, y), and B x,y is its value after binarisation ISODATA method Thresholding using ISIDATA [10] consists to find a threshold by separating iteratively the gray-level histogram into two classes, with the apriority knowledge of the values associate to each class. This method starts by dividing the interval of non-null values into two equidistant parts, and next we take m 1 and m 2 as the arithmetic means of each class. Repeat until convergence, the calculation of the threshold T as the closest integer to (m 1 +m 2 )/2 and update the two means m 1 and m Kapur method The kapur method [11] is considered as an extension of the Pun method [20]. In contrast to this last, the Kapur method takes in the account the object likelihood distribution P t and the background likelihood distribution (1- P t ) in the determination of the division entropy. The binarisation threshold T is chosen of which the value: H=H object +H background be maximal. T is given thus by the expression: T = max (H object + H background ) = max ( t P i P t i=0 log P i P t 255 i=t+1 P i 1 P t log P i 1 P t ) Such as P i is the likelihood of occurrence of the gray-level i in the image, and P t = t i=0 P i Cheng-Chen method The Cheng-Chen method [12] is based on the maximal entropy principle and the fizzy C-partition, for the calculation of the threshold. Consider two fuzzy sets Object and Background of which the membership functions are defined by: 1, x a µ object = (x-c)/ (a-c) a<x<c 0, x c 0, x a µ background = (x-a)/(c-a) a<x<c 1, x c In this method, the binarization threshold is chosen as the gray-level of which the membership function=0.5, and so it is the center of the interval [a opt, c opt ], such as a opt and c opt are the values of a and c maximizing the division entropy. Cheng and Chen proposed an algorithm for finding the optimal values of a and c corresponding to the maximal partitioning entropy. The partitioning entropy is given by: H= -P(object) log (P(object))-P(background) log (P(background)) Li-Lee method Li and Lee [13] proposed an images thresholding technique where the grouping of gray-levels in two classes is based on the crossed entropy minimization. Note µ 1 (t) and µ 2 (t) the means of the two classes according to the gray-level t, and h i the histogram value at the gray-level i. Li and Lee established that the optimal threshold T is calculated in order to minimize η(t)= η 1 (t)+ η 2 (t) where: 2.2. Local methods η 1 t = t 1 i=1 i i h i log ( μ 1 (t) ) and η 2 t = μ 1 t = t 1 i h i i=1 t 1 i=1 h i et μ 2 t = 256 i=t i i h i log ( μ 2 (t) ) 256 i h i Bernsen method It is an adaptive local method of which the threshold is calculated for each pixel of the image [14]. Thus for each pixel of coordinates (x, y), the threshold is given by: i=t 256 i=t h i

4 T x, y = Z low + Z high 2 Such as Z low and Z high are the lowest and the highest gray-level respectively in a squared window r r centered over the pixel (x, y). However if the contrast measurement C(x, y)= (Z haut - Z bas ) is lower than a threshold l, then the neighborhood consists of a single class: background or text. In [22], the tests performed shows that the choice of l=15 and r=15gives the bests results Niblack method The Niblack algorithm [15] calculates a local threshold for each pixel by gliding a rectangular window over the whole image. The threshold T is calculated by using the mean m and the standard deviation σ of all pixels in the window (neighborhood of the pixel in question). Thus the threshold T is given by: T = m + k* σ Such as k is a parameter used for determining the number of edge pixels considered as object pixels, and takes a negative values (k is fixed -0.2 by authors). In [21], a window of size gives good results. There is in the literature several evaluation works of binarization methods of which several between them [22] [1], consider the Niblack method (together with a post-treatment step of Yanowitz and Bruckstein [23]) as the better Sauvola method The Sauvola algorithm [16] is a modification of that of Niblack, in order to gives more performance in the documents with a background containing a light texture or too variation and uneven illumination. In the Sauvola modification, the binarization is given by: T = m. (1 k. (1 σ R )) Where R is the dynamic carried of the standard deviation σ, and the parameter k takes positives values in the interval [0.2, 0.5]. In his tests, Sauvola used R=128 and k=0.5 According to [6], the Sauvola method is shown more effective that Niblack method when the gray-level of the text is close to 0 and that of background is close to 255. However in images where the gray-level of background and text pixels are closes, the results are unsatisfactory Wolf method In order to address to Sauvola algorithm problems (low contrast, disparity of gray-level, etc.), Wolf and al. [17] proposed to normalize the contrast and the image gray-level mean, and calculate the threshold by: T = 1 k m + k M + k σ (m M) R Such as k is fixed to 0.5, M is the minimum gray-level of the image and R is the maximum standard deviation of gray-levels obtained over all windows Nick method The major advantage of this method is that it improves considerably the binarization of white pages and lighted images and in the case where the image presents low contrast, by downwards moving, the threshold of binarization [6]. The threshold T is calculated as follow:

5 ( p 2 T = m + k i m 2 ) NP Such as: k is the Niblack factor and vary between -0.1 and -0.2 according to the application need, m: the average gray-level, p i : the gray-level of pixel i and NP is the total number of pixels. Kurshid and al assume that this method works very well for the most (if not all) type of degraded old documents. In order to evaluate their performances, they applied the method globally and locally over images documents. The window size in the case of local test is chosen equal to Fuzzy hierarchical segmentation It is an adaptive thresholding method based on hierarchical principle to palliate the non-uniform illumination problem presented in the most of real images [18]. This method consists of two steps: Hierarchy construction In this step we construct the quaternary tree, by decomposing consecutively the image into four sub-images by evaluating a homogeneity criterion. Each sub-image is associated to a node of the image. If one sub-image satisfied the homogeneity criterion, the division is not necessary, and the sub-image becomes a terminal node of the image. In the opposed case, this sub-image is decomposed into four parts. The process continues until all the nodes in the tree become terminals Decision and defuzzification For each hierarchical level, we calculate the membership degree of each pixel (m, n) to the Object and Background classes. We use for that, the Zadeh S function. The membership degree is given by: µ (m, n) = S (x mn ; m σ, m, m + σ) Such as: m and σ are respectively the mean and the standard deviation of the given sub-image gray-level and x mn is the gray-level of the pixel (m, n). Afterwards, we use the t-conorme Zadeh operator [24] to calculate the final membership degree of each pixel (m, n) to the Object (µ F o (m, n)) and Background classes (µ F f (m, n)). A pixel is affected thus to the Object class if its membership degree incorporated to the Object class is upper than that incorporated to the Background class. Else it is affected to the Background class. 3. Experimentation and results In this section, we present at first our test collection on which we applied the binarization methods above mentioned and next the obtained results Test images We collected a database of Arabic manuscript old images grace to different partnership projects realized by the two laboratories of our department that is the LRI and the LabGed. These document images are of different nature, having several characteristics coming from their natures, their storage state: degradation characteristics (stains, holes, humidity traces, ink degradation and thin lines, transparence effect, presence of folds and tears etc.), characteristics attached to documents nature ( the page may contain graphic areas of different sizes, low separation between text blocks, high variability between writing styles, irregular inter-words space, variations of paper color etc.), and characteristics of failing digitization ( of curvature, of light etc.) All these characteristics make difficult all type off processing and analysis of these images. Figure 1 shows some Arabic manuscript old images of our test collection having a high complexity degree.

6 Figure 1: Some examples of documents images of our test collection 3.2. Obtained results We applied the binarization methods previously presented over our test collection composed of 120 old Arabic documents images having several types of degradations and structure complexity. We present in this section the results of applying the previous methods over three images of our collection. For local methods, we tested several values of parameters k, R and the size r of the sliding window, and we choose those which give the best results. So for the Bernsen method we fixed r=15 15, for Niblack method, k = -0.2 and r=25 25, for Sauvola method, k= 0.2, R= 128 and r= For Wolf method the values k= 0.2 and r= 7 7 are chosen and for Nick method, k=-0.2 and r= Originals images Global methods Absolute S=127 Otsu

7 Otsu 2 thresholds ISODATA Kapur Cheng and Chen Li and Lee Nick k= -0.1

8 Hybrids methods Otsu ISODATA Locales methods (r designate the window size) Bernsen r=15 15 Niblack k= -0.2 r=25 25 Sauvola k= 0.2 R= 128 r= Wolf k= 0.2 r= 7 7

9 Nick k=-0.2 r=19 19 Gadi and Benslimane 3.3. Evaluation In several works, the evaluation of document images binarization methods is based on the calculation of the recognition rate of characters presents in the documents [22] [6] [21], by using generally a commercial OCR software as for example ABBYY Fine Reader. In other works, the evaluation is performed by calculating a distance (edit-distance for example) between the images resulting from binarization and Ground truth images [7] [3] [6]. Applying the OCR as an evaluation criterion, is not possible in our case, because the absence of an efficient commercial software for recognizing Arabic manuscript writing. And as Ground truth images are not available for our collection, using the distance with the images resulting from binarization is impossible too. As we previously said, the goal of our comparative study is to choose the best binarisation algorithm to use it in our Arabic documents retrieval system. We proposed thus to use this system in the comparison of binarisation methods. For that, we firstly assigned to each image of our test collection a textual signature (code) manually. This signature represents structural features of the writing present in the image (descenders, ascenders, loops and diacritics), and it is obtained by corresponding to each feature a precise code. We consider these signatures as Ground truth signatures. Over images resulting from the different binarization methods, we apply the series of processing proposed in our Arabic documents retrieval system [26] (smoothing, line and sub-word segmentation etc.) finishing by an extraction of structural features of characters presents in documents. Features extracted from each image are coded in ASCII and make up the signature of this image. For each image of our test collection, we compare the corresponding ground truth signature with signatures extracted from images binarized by using the different binarization methods, and we use for that the famous editdistance [27]. The best binarization algorithm for each image is thus that with which the obtained signature is the most close to the ground truth signature. The Table 1 recapitulates averages distances obtained between signatures extracted from images binarized by using the different binarization methods and the ground truth signatures.

10 Method Average distance Fixed global thresholding Otsu Otsu 2 seuils ISODATA 39 Kapur 64 Cheng and Chen 227 Li-Lee Global Nick Otsu hybride ISODATA hybride Bernsen Niblack 146 Sauvola Wolf Local Nick Fuzzy hierarchical segmentation Table 1: averages distances obtained for the different binarization methods From the precedent table, it is shown that the Nick method of Khirshid and al. [6] is the best for our test collection. 4. Conclusion In this paper, we tackled the general problem of binarizing degraded old Arabic documents. A lot of methods were proposed in the literature for binarizing documents images, but no one between them demarcate to the others, especially for the old documents images having different degradation characteristics. We presented a little state of the art over documents images thresholding techniques the most frequently cited in the literature. An implementation of these methods is performed by applying them over a set of 120 old Arabic documents images containing different types of degradations, and structure complexity. Because the absence of an efficient commercial software for recognizing Arabic manuscript writing, and the no- availability of ground truth images for our collection, we propose to use a new evaluation method. This last is based on the comparison of features extracted from images binarized by using the different binarization methods with ground truth features. The best binarization method is that with which the obtained features are the most closes to the ground truth features. The results showed that the Nick method is the best. 5. References [1] Ø. D. Trier and T. Taxt, «Evaluation of binarization methods for document images», IEEE Transactions on Pattern Analysis and Machine Intelligence, 17(3), p , March [2] G. Leedham, S. Varma, A. Patankar, and V. Govindaraju, «Separating text and background in degraded documents images - A comparison of global thresholding techniques for multi-stage thresholding», Proceedings of the 8 th International Workshop on Frontiers in Handwriting Recognition, p , Canada, August [3] J. Sauvola, M. Pietikainen, «Adaptive document image binarization», Pattern Recognition, 33(2), p , [4] C. M. Tsai and H. J. Lee, «Binarization of color document images via luminance and saturation color features», IEEE Transactions on Image Processing, 11, p , April [5] N. Arica, F. T. Yarman-Vural, «An overview of character recognition focused on off-line handwriting», IEEE transactions on systems, man and cybernetics - part C : Applications and reviews, 31(2), p , [6] K. Khurshid, I. Siddiqi, C. Faure, N. Vincent, «Comparison of Niblack inspired Binarization methods for ancient documents», 16 th International conference on Document Recognition and Retrieval, USA, [7] M. Sezgin, B. Sankur, «Survey over image thresholding techniques and quantitative performance evaluation», Journal of Electronic Imaging, 13(1), p , January 2004.

11 [8] M. R. Gupta, N. P. Jacobson, E. K. Garcia, «OCR binarization and image pre-processing for searching historical documents», Pattern Recognition, 40, p , [9] N. OTSU, «A Threshold Selection Method from Gray-Level Histograms», IEEE transactions on Systems, Man and Cybernetics, 9(1), p , [10] F.R.D. Velasco, «Thresholding using the ISODATA clustering algorithm», IEEE Transaction on system, Man and Cybernitics, 10, p , [11] J.N. Kapur, P.K. Sahoo, A.K.C. Wong, «A New method for gray-level picture threshold using the entropy of the histogram», Computer Vision, Graphics, and Image Processing, 29, p , [12] H.D. Cheng, J. R. Chen and J. Li, «Threshold selection based on fuzzy c-partition entropy approach», Pattern Recognition, 31(7), p , [13] C.H. Li, C.K. Lee, «Minimum Cross Entropy Thresholding», Pattern Recognition, 26(4), p , [14] J. Bernsen, «Dynamic thresholding of grey-level images», in Proc. 8 th International Conference on Pattern Recognition, p , Paris, France, [15] W. Niblack, «An Introduction to Digital Image Processing», Ed. Prentice Hall, Englewood Cliffs, [16] J. Sauvola, T. Seppänen, S. Haapakoski, M. Pietikänen, «Adaptive document binarisation», Fourth International Conference Document Analysis and Recognition (ICDAR), p , Ulm, Germany, August [17] C. Wolf, J. M. Jolion, and F. Chassaing. «Extraction de texte dans des vidéos : le cas de la binarisation», In 13 ème Congrès Francophone de Reconnaissance des Formes et Intelligence Artificielle, p , [18] T. Gadi, R. Benslimane, «Segmentation hiérarchique floue», Traitement du signal, 7(1), [19] C. Wolf, J. M. Jolion, «Détection et extraction de texte de la vidéo», CFED, p , [20] T. Pun, «A New method for gray-level picture threshold using the entropy of the histogram», Signal processing, 2(3), p , [21] J. He, Q. D. M. Do, A. C. Downton, and J. H. Kim, «A comparison of binarization methods for historical archive documents», International Conference on Document Analysis and Recognition (ICDAR), p , [22] A.J.O. Trier, «Goal-directed evaluation of binarization methods», IEEE Transactions on Pattern Analysis and Machine Intelligence, 17(12), p , [23] S.D. Yanomitz and A.M. Bruckstein, «A new method for image segmentation», In Computer Vision, Graphics and Image Processing, 16(1), p , [24] L. A. Zadeh, «Fuzzy Sets and their Applications to cognitive and Decision Processes», p. 1-39, Ed. Academic Press, London, [25] L. O Gorman, «Experimental comparisons of binarization and multithresholding methods on document images», in: Proceedings of the IAPR International Conference on Pattern Recognition, vol. 2, IEEE, 1994, pp [26] T. Sari, A. Kefali, «A search engine for Arabic documents», Actes du dixième Colloque International Francophone sur l Écrit et le Document (CIFED), p , [27] Levenstein A., «Binary Codes Capable of Correcting Deletions, Insertions and Reversals». Sov.Phy. Dohl. Vol.10, P , 1966

Text extraction from historical document images by the combination of several thresholding techniques

Text extraction from historical document images by the combination of several thresholding techniques Hindawi Publishing Corporation Advances in Multimedia Research article Text extraction from historical document images by the combination of several thresholding techniques Abderrahmane KEFALI 1, Toufik

More information

An Objective Evaluation Methodology for Handwritten Image Document Binarization Techniques

An Objective Evaluation Methodology for Handwritten Image Document Binarization Techniques An Objective Evaluation Methodology for Handwritten Image Document Binarization Techniques K. Ntirogiannis, B. Gatos and I. Pratikakis Computational Intelligence Laboratory, Institute of Informatics and

More information

A Novel q-parameter Automation in Tsallis Entropy for Image Segmentation

A Novel q-parameter Automation in Tsallis Entropy for Image Segmentation A Novel q-parameter Automation in Tsallis Entropy for Image Segmentation M Seetharama Prasad KL University Vijayawada- 522202 P Radha Krishna KL University Vijayawada- 522202 ABSTRACT Image Thresholding

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

EUSIPCO

EUSIPCO EUSIPCO 2013 1569743917 BINARIZATION OF HISTORICAL DOCUMENTS USING SELF-LEARNING CLASSIFIER BASED ON K-MEANS AND SVM Amina Djema and Youcef Chibani Speech Communication and Signal Processing Laboratory

More information

SEPARATING TEXT AND BACKGROUND IN DEGRADED DOCUMENT IMAGES A COMPARISON OF GLOBAL THRESHOLDING TECHNIQUES FOR MULTI-STAGE THRESHOLDING

SEPARATING TEXT AND BACKGROUND IN DEGRADED DOCUMENT IMAGES A COMPARISON OF GLOBAL THRESHOLDING TECHNIQUES FOR MULTI-STAGE THRESHOLDING SEPARATING TEXT AND BACKGROUND IN DEGRADED DOCUMENT IMAGES A COMPARISON OF GLOBAL THRESHOLDING TECHNIQUES FOR MULTI-STAGE THRESHOLDING GRAHAM LEEDHAM 1, SAKET VARMA 2, ANISH PATANKAR 2 and VENU GOVINDARAJU

More information

ESTIMATION OF APPROPRIATE PARAMETER VALUES FOR DOCUMENT BINARIZATION TECHNIQUES

ESTIMATION OF APPROPRIATE PARAMETER VALUES FOR DOCUMENT BINARIZATION TECHNIQUES International Journal of Robotics and Automation, Vol. 24, No. 1, 2009 ESTIMATION OF APPROPRIATE PARAMETER VALUES FOR DOCUMENT BINARIZATION TECHNIQUES E. Badekas and N. Papamarkos Abstract Most of the

More information

arxiv:cs/ v1 [cs.cv] 12 Feb 2006

arxiv:cs/ v1 [cs.cv] 12 Feb 2006 Multilevel Thresholding for Image Segmentation through a Fast Statistical Recursive Algorithm arxiv:cs/0602044v1 [cs.cv] 12 Feb 2006 S. Arora a, J. Acharya b, A. Verma c, Prasanta K. Panigrahi c 1 a Dhirubhai

More information

Random spatial sampling and majority voting based image thresholding

Random spatial sampling and majority voting based image thresholding 1 Random spatial sampling and majority voting based image thresholding Yi Hong Y. Hong is with the City University of Hong Kong. yihong@cityu.edu.hk November 1, 7 2 Abstract This paper presents a novel

More information

IJSER. Abstract : Image binarization is the process of separation of image pixel values as background and as a foreground. We

IJSER. Abstract : Image binarization is the process of separation of image pixel values as background and as a foreground. We International Journal of Scientific & Engineering Research, Volume 7, Issue 3, March-2016 1238 Adaptive Local Image Contrast in Image Binarization Prof.Sushilkumar N Holambe. PG Coordinator ME(CSE),College

More information

Binarization of Degraded Historical Document Images

Binarization of Degraded Historical Document Images Binarization of Degraded Historical Document Images Zineb Hadjadj Université de Blida Blida, Algérie hadjadj_zineb@yahoo.fr Mohamed Cheriet École de Technologie Supérieure Montréal, Canada mohamed.cheriet@etsmtl.ca

More information

Multi-pass approach to adaptive thresholding based image segmentation

Multi-pass approach to adaptive thresholding based image segmentation 1 Multi-pass approach to adaptive thresholding based image segmentation Abstract - Thresholding is still one of the most common approaches to monochrome image segmentation. It often provides sufficient

More information

Local Contrast and Mean based Thresholding Technique in Image Binarization

Local Contrast and Mean based Thresholding Technique in Image Binarization Local Contrast and Mean based Thresholding Technique in Image Binarization O. Imocha Singh Department of Computer Science, Manipur University Canchipur, Imphal-795003, Manipur, India. O. James DOEACC,

More information

Document Image Binarization Using Post Processing Method

Document Image Binarization Using Post Processing Method Document Image Binarization Using Post Processing Method E. Balamurugan Department of Computer Applications Sathyamangalam, Tamilnadu, India E-mail: rethinbs@gmail.com K. Sangeetha Department of Computer

More information

ADAPTIVE DOCUMENT IMAGE THRESHOLDING USING IFOREGROUND AND BACKGROUND CLUSTERING

ADAPTIVE DOCUMENT IMAGE THRESHOLDING USING IFOREGROUND AND BACKGROUND CLUSTERING ADAPTIVE DOCUMENT IMAGE THRESHOLDING USING IFOREGROUND AND BACKGROUND CLUSTERING Andreas E. Savakis Eastman Kodak Company 901 Elmgrove Rd. Rochester, New York 14653 savakis @ kodak.com ABSTRACT Two algorithms

More information

A proposed optimum threshold level for document image binarization

A proposed optimum threshold level for document image binarization 7, Issue 1 (2017) 8-14 Journal of Advanced Research in Computing and Applications Journal homepage: www.akademiabaru.com/arca.html ISSN: 2462-1927 A proposed optimum threshold level for document image

More information

SSRG International Journal of Computer Science and Engineering (SSRG-IJCSE) volume1 issue7 September 2014

SSRG International Journal of Computer Science and Engineering (SSRG-IJCSE) volume1 issue7 September 2014 SSRG International Journal of Computer Science and Engineering (SSRG-IJCSE) volume issue7 September 24 A Thresholding Method for Color Image Binarization Kalavathi P Department of Computer Science and

More information

Multicriteria Image Thresholding Based on Multiobjective Particle Swarm Optimization

Multicriteria Image Thresholding Based on Multiobjective Particle Swarm Optimization Applied Mathematical Sciences, Vol. 8, 2014, no. 3, 131-137 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ams.2014.3138 Multicriteria Image Thresholding Based on Multiobjective Particle Swarm

More information

SCIENCE & TECHNOLOGY

SCIENCE & TECHNOLOGY Pertanika J. Sci. & Technol. 25 (S): 63-72 (2017) SCIENCE & TECHNOLOGY Journal homepage: http://www.pertanika.upm.edu.my/ Binarization via the Dynamic Histogram and Window Tracking for Degraded Textual

More information

RESTORATION OF DEGRADED DOCUMENTS USING IMAGE BINARIZATION TECHNIQUE

RESTORATION OF DEGRADED DOCUMENTS USING IMAGE BINARIZATION TECHNIQUE RESTORATION OF DEGRADED DOCUMENTS USING IMAGE BINARIZATION TECHNIQUE K. Kaviya Selvi 1 and R. S. Sabeenian 2 1 Department of Electronics and Communication Engineering, Communication Systems, Sona College

More information

Historical Handwritten Document Image Segmentation Using Background Light Intensity Normalization

Historical Handwritten Document Image Segmentation Using Background Light Intensity Normalization Historical Handwritten Document Image Segmentation Using Background Light Intensity Normalization Zhixin Shi and Venu Govindaraju Center of Excellence for Document Analysis and Recognition (CEDAR), State

More information

A NOVEL BINARIZATION METHOD FOR QR-CODES UNDER ILL ILLUMINATED LIGHTINGS

A NOVEL BINARIZATION METHOD FOR QR-CODES UNDER ILL ILLUMINATED LIGHTINGS A NOVEL BINARIZATION METHOD FOR QR-CODES UNDER ILL ILLUMINATED LIGHTINGS N.POOMPAVAI 1 Dr. R. BALASUBRAMANIAN 2 1 Research Scholar, JJ College of Arts and Science, Bharathidasan University, Trichy. 2 Research

More information

CHAPTER II LITERATURE SURVEY

CHAPTER II LITERATURE SURVEY CHAPTER II LITERATURE SURVEY 2.1 Introduction An image is considered as a set of points or pixels distributed over a two dimensional finite space. Segmentation, in precise form, can be considered as the

More information

New Binarization Approach Based on Text Block Extraction

New Binarization Approach Based on Text Block Extraction 2011 International Conference on Document Analysis and Recognition New Binarization Approach Based on Text Block Extraction Ines Ben Messaoud, Hamid Amiri Laboratoire des Systèmes et Traitement de Signal

More information

Evaluation of document binarization using eigen value decomposition

Evaluation of document binarization using eigen value decomposition Evaluation of document binarization using eigen value decomposition Deepak Kumar, M N Anil Prasad and A G Ramakrishnan Medical Intelligence and Language Engineering Laboratory Department of Electrical

More information

A Function-Based Image Binarization based on Histogram

A Function-Based Image Binarization based on Histogram International Research Journal of Applied and Basic Sciences 2015 Available online at www.irjabs.com ISSN 2251-838X / Vol, 9 (3): 418-426 Science Explorer Publications A Function-Based Image Binarization

More information

Advances in Natural and Applied Sciences. Efficient Illumination Correction for Camera Captured Image Documents

Advances in Natural and Applied Sciences. Efficient Illumination Correction for Camera Captured Image Documents AENSI Journals Advances in Natural and Applied Sciences ISSN:1995-0772 EISSN: 1998-1090 Journal home page: www.aensiweb.com/anas Efficient Illumination Correction for Camera Captured Image Documents 1

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

Adaptive Local Thresholding for Fluorescence Cell Micrographs

Adaptive Local Thresholding for Fluorescence Cell Micrographs TR-IIS-09-008 Adaptive Local Thresholding for Fluorescence Cell Micrographs Jyh-Ying Peng and Chun-Nan Hsu November 11, 2009 Technical Report No. TR-IIS-09-008 http://www.iis.sinica.edu.tw/page/library/lib/techreport/tr2009/tr09.html

More information

Small-scale objects extraction in digital images

Small-scale objects extraction in digital images 102 Int'l Conf. IP, Comp. Vision, and Pattern Recognition IPCV'15 Small-scale objects extraction in digital images V. Volkov 1,2 S. Bobylev 1 1 Radioengineering Dept., The Bonch-Bruevich State Telecommunications

More information

arxiv:cs/ v1 [cs.cv] 9 Mar 2006

arxiv:cs/ v1 [cs.cv] 9 Mar 2006 Locally Adaptive Block Thresholding Method with Continuity Constraint arxiv:cs/0603041v1 [cs.cv] 9 Mar 2006 S. Hemachander a, A. Verma b, S. Arora c, Prasanta K. Panigrahi b1 a University of Buffalo, NY,

More information

Package autothresholdr

Package autothresholdr Type Package Package autothresholdr January 23, 2018 Title An R Port of the 'ImageJ' Plugin 'Auto Threshold' Version 1.1.0 Maintainer Rory Nolan Provides the 'ImageJ' 'Auto Threshold'

More information

Document image binarisation using Markov Field Model

Document image binarisation using Markov Field Model 009 10th International Conference on Document Analysis and Recognition Document image binarisation using Markov Field Model Thibault Lelore, Frédéric Bouchara UMR CNRS 6168 LSIS Southern University of

More information

A Clustering-Based Method for. Brain Tumor Segmentation

A Clustering-Based Method for. Brain Tumor Segmentation Contemporary Engineering Sciences, Vol. 9, 2016, no. 15, 743-754 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ces.2016.6564 A Clustering-Based Method for Brain Tumor Segmentation Idanis Diaz

More information

Binarization of Document Images: A Comprehensive Review

Binarization of Document Images: A Comprehensive Review Journal of Physics: Conference Series PAPER OPEN ACCESS Binarization of Document Images: A Comprehensive Review To cite this article: Wan Azani Mustafa and Mohamed Mydin M. Abdul Kader 2018 J. Phys.: Conf.

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

Digital Image Processing. Lecture # 15 Image Segmentation & Texture

Digital Image Processing. Lecture # 15 Image Segmentation & Texture Digital Image Processing Lecture # 15 Image Segmentation & Texture 1 Image Segmentation Image Segmentation Group similar components (such as, pixels in an image, image frames in a video) Applications:

More information

Effective Features of Remote Sensing Image Classification Using Interactive Adaptive Thresholding Method

Effective Features of Remote Sensing Image Classification Using Interactive Adaptive Thresholding Method Effective Features of Remote Sensing Image Classification Using Interactive Adaptive Thresholding Method T. Balaji 1, M. Sumathi 2 1 Assistant Professor, Dept. of Computer Science, Govt. Arts College,

More information

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

A derivative based algorithm for image thresholding

A derivative based algorithm for image thresholding A derivative based algorithm for image thresholding André Ricardo Backes arbackes@yahoo.com.br Bruno Augusto Nassif Travençolo travencolo@gmail.com Mauricio Cunha Escarpinati escarpinati@gmail.com Faculdade

More information

Segmentation of Arabic handwritten text to lines

Segmentation of Arabic handwritten text to lines Available online at www.sciencedirect.com ScienceDirect Procedia Computer Science 73 (2015 ) 115 121 The International Conference on Advanced Wireless, Information, and Communication Technologies (AWICT

More information

Input sensitive thresholding for ancient Hebrew manuscript

Input sensitive thresholding for ancient Hebrew manuscript Pattern Recognition Letters 26 (2005) 1168 1173 www.elsevier.com/locate/patrec Input sensitive thresholding for ancient Hebrew manuscript Itay Bar-Yosef * Department of Computer Science, Ben Gurion University,

More information

A Review of Evaluation of Optimal Binarization Technique for Character Segmentation in Historical Manuscripts

A Review of Evaluation of Optimal Binarization Technique for Character Segmentation in Historical Manuscripts 010 Third International Conerence on Knowledge Discovery and Data Mining A Review o Evaluation o Optimal Binarization Technique or Character Segmentation in Historical Manuscripts Chun Che Fung and Rapeeporn

More information

A Statistical approach to line segmentation in handwritten documents

A Statistical approach to line segmentation in handwritten documents A Statistical approach to line segmentation in handwritten documents Manivannan Arivazhagan, Harish Srinivasan and Sargur Srihari Center of Excellence for Document Analysis and Recognition (CEDAR) University

More information

Best Combination of Binarization Methods for License Plate Character Segmentation

Best Combination of Binarization Methods for License Plate Character Segmentation Best Combination of Binarization Methods for License Plate Character Segmentation Youngwoo Yoon, Kyu-Dae Ban, Hosub Yoon, Jaeyeon Lee, and Jaehong Kim A connected component analysis from a binary image

More information

Image restoration. Restoration: Enhancement:

Image restoration. Restoration: Enhancement: Image restoration Most images obtained by optical, electronic, or electro-optic means is likely to be degraded. The degradation can be due to camera misfocus, relative motion between camera and object,

More information

MURDOCH RESEARCH REPOSITORY.

MURDOCH RESEARCH REPOSITORY. MURDOCH RESEARCH REPOSITORY http://researchrepository.murdoch.edu.au This is the author's final version of the work, as accepted for publication following peer review but without the publisher's layout

More information

A Quantitative Approach for Textural Image Segmentation with Median Filter

A Quantitative Approach for Textural Image Segmentation with Median Filter International Journal of Advancements in Research & Technology, Volume 2, Issue 4, April-2013 1 179 A Quantitative Approach for Textural Image Segmentation with Median Filter Dr. D. Pugazhenthi 1, Priya

More information

Multi-scale Techniques for Document Page Segmentation

Multi-scale Techniques for Document Page Segmentation Multi-scale Techniques for Document Page Segmentation Zhixin Shi and Venu Govindaraju Center of Excellence for Document Analysis and Recognition (CEDAR), State University of New York at Buffalo, Amherst

More information

A New iterative triclass thresholding technique for Image Segmentation

A New iterative triclass thresholding technique for Image Segmentation A New iterative triclass thresholding technique for Image Segmentation M.M.Raghavendra Asst Prof, Department of ECE Brindavan Institute of Technology & Science Kurnool, India E-mail: mmraghavendraece@gmail.com

More information

An Efficient Character Segmentation Based on VNP Algorithm

An Efficient Character Segmentation Based on VNP Algorithm Research Journal of Applied Sciences, Engineering and Technology 4(24): 5438-5442, 2012 ISSN: 2040-7467 Maxwell Scientific organization, 2012 Submitted: March 18, 2012 Accepted: April 14, 2012 Published:

More information

Robust Phase-Based Features Extracted From Image By A Binarization Technique

Robust Phase-Based Features Extracted From Image By A Binarization Technique IOSR Journal of Computer Engineering (IOSR-JCE) e-issn: 2278-0661,p-ISSN: 2278-8727, Volume 18, Issue 4, Ver. IV (Jul.-Aug. 2016), PP 10-14 www.iosrjournals.org Robust Phase-Based Features Extracted From

More information

Short Survey on Static Hand Gesture Recognition

Short Survey on Static Hand Gesture Recognition Short Survey on Static Hand Gesture Recognition Huu-Hung Huynh University of Science and Technology The University of Danang, Vietnam Duc-Hoang Vo University of Science and Technology The University of

More information

Learning to Segment Document Images

Learning to Segment Document Images Learning to Segment Document Images K.S. Sesh Kumar, Anoop Namboodiri, and C.V. Jawahar Centre for Visual Information Technology, International Institute of Information Technology, Hyderabad, India Abstract.

More information

Handwritten Script Recognition at Block Level

Handwritten Script Recognition at Block Level Chapter 4 Handwritten Script Recognition at Block Level -------------------------------------------------------------------------------------------------------------------------- Optical character recognition

More information

Final Exam Schedule. Final exam has been scheduled. 12:30 pm 3:00 pm, May 7. Location: INNOVA It will cover all the topics discussed in class

Final Exam Schedule. Final exam has been scheduled. 12:30 pm 3:00 pm, May 7. Location: INNOVA It will cover all the topics discussed in class Final Exam Schedule Final exam has been scheduled 12:30 pm 3:00 pm, May 7 Location: INNOVA 1400 It will cover all the topics discussed in class One page double-sided cheat sheet is allowed A calculator

More information

Texture Analysis of Painted Strokes 1) Martin Lettner, Paul Kammerer, Robert Sablatnig

Texture Analysis of Painted Strokes 1) Martin Lettner, Paul Kammerer, Robert Sablatnig Texture Analysis of Painted Strokes 1) Martin Lettner, Paul Kammerer, Robert Sablatnig Vienna University of Technology, Institute of Computer Aided Automation, Pattern Recognition and Image Processing

More information

Image Segmentation Based on Watershed and Edge Detection Techniques

Image Segmentation Based on Watershed and Edge Detection Techniques 0 The International Arab Journal of Information Technology, Vol., No., April 00 Image Segmentation Based on Watershed and Edge Detection Techniques Nassir Salman Computer Science Department, Zarqa Private

More information

Haar-like-features for query-by-string word spotting

Haar-like-features for query-by-string word spotting Haar-like-features for query-by-string word spotting Adam Ghorbel, Jean-Marc Ogier, Nicole Vincent To cite this version: Adam Ghorbel, Jean-Marc Ogier, Nicole Vincent. Haar-like-features for query-by-string

More information

CS 490: Computer Vision Image Segmentation: Thresholding. Fall 2015 Dr. Michael J. Reale

CS 490: Computer Vision Image Segmentation: Thresholding. Fall 2015 Dr. Michael J. Reale CS 490: Computer Vision Image Segmentation: Thresholding Fall 205 Dr. Michael J. Reale FUNDAMENTALS Introduction Before we talked about edge-based segmentation Now, we will discuss a form of regionbased

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

Quality evaluation of degraded document images for binarization result prediction

Quality evaluation of degraded document images for binarization result prediction Quality evaluation of degraded document images for binarization result prediction Vincent Rabeux, Nicholas Journet, Anne Vialard, Jean-Philippe Domenger To cite this version: Vincent Rabeux, Nicholas Journet,

More information

Foreground-Background Separation by Feed-forward Neural Networks in Old Manuscripts

Foreground-Background Separation by Feed-forward Neural Networks in Old Manuscripts Informatica 38 (014) 39 338 39 Foreground-Background Separation by Feed-forward Neural Networks in Old Manuscripts Abderrahmane Kefali, Toufik Sari and Halima Bahi LabGED Laboratory Computer Science Department,

More information

A Weighting Scheme for Improving Otsu Method for Threshold Selection

A Weighting Scheme for Improving Otsu Method for Threshold Selection A Weighting Scheme for Improving Otsu Method for Threshold Selection Hui-Fuang Ng, Cheng-Wai Kheng, and Jim-Min in Department of Computer Science, Universiti Tunku Abdul Rahman, Kampar 3900, Perak, Malaysia

More information

Image segmentation. Stefano Ferrari. Università degli Studi di Milano Methods for Image Processing. academic year

Image segmentation. Stefano Ferrari. Università degli Studi di Milano Methods for Image Processing. academic year Image segmentation Stefano Ferrari Università degli Studi di Milano stefano.ferrari@unimi.it Methods for Image Processing academic year 2017 2018 Segmentation by thresholding Thresholding is the simplest

More information

Image Enhancement: To improve the quality of images

Image Enhancement: To improve the quality of images Image Enhancement: To improve the quality of images Examples: Noise reduction (to improve SNR or subjective quality) Change contrast, brightness, color etc. Image smoothing Image sharpening Modify image

More information

IMPLEMENTATION OF SPATIAL FUZZY CLUSTERING IN DETECTING LIP ON COLOR IMAGES

IMPLEMENTATION OF SPATIAL FUZZY CLUSTERING IN DETECTING LIP ON COLOR IMAGES IMPLEMENTATION OF SPATIAL FUZZY CLUSTERING IN DETECTING LIP ON COLOR IMAGES Agus Zainal Arifin 1, Adhatus Sholichah 2, Anny Yuniarti 3, Dini Adni Navastara 4, Wijayanti Nurul Khotimah 5 1,2,3,4,5 Department

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

Effect of Pre-Processing on Binarization

Effect of Pre-Processing on Binarization Boise State University ScholarWorks Electrical and Computer Engineering Faculty Publications and Presentations Department of Electrical and Computer Engineering 1-1-010 Effect of Pre-Processing on Binarization

More information

Efficient binarization technique for severely degraded document images

Efficient binarization technique for severely degraded document images CSIT (November 2014) 2(3):153 161 DOI 10.1007/s40012-014-0045-5 ORIGINAL ARTICLE Efficient binarization technique for severely degraded document images Brij Mohan Singh Mridula Received: 22 April 2014

More information

Optical Character Recognition (OCR) for Printed Devnagari Script Using Artificial Neural Network

Optical Character Recognition (OCR) for Printed Devnagari Script Using Artificial Neural Network International Journal of Computer Science & Communication Vol. 1, No. 1, January-June 2010, pp. 91-95 Optical Character Recognition (OCR) for Printed Devnagari Script Using Artificial Neural Network Raghuraj

More information

Enhanced Image. Improved Dam point Labelling

Enhanced Image. Improved Dam point Labelling 3rd International Conference on Multimedia Technology(ICMT 2013) Video Text Extraction Based on Stroke Width and Color Xiaodong Huang, 1 Qin Wang, Kehua Liu, Lishang Zhu Abstract. Video text can be used

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

AN EFFICIENT BINARIZATION TECHNIQUE FOR FINGERPRINT IMAGES S. B. SRIDEVI M.Tech., Department of ECE

AN EFFICIENT BINARIZATION TECHNIQUE FOR FINGERPRINT IMAGES S. B. SRIDEVI M.Tech., Department of ECE AN EFFICIENT BINARIZATION TECHNIQUE FOR FINGERPRINT IMAGES S. B. SRIDEVI M.Tech., Department of ECE sbsridevi89@gmail.com 287 ABSTRACT Fingerprint identification is the most prominent method of biometric

More information

Two Algorithms of Image Segmentation and Measurement Method of Particle s Parameters

Two Algorithms of Image Segmentation and Measurement Method of Particle s Parameters Appl. Math. Inf. Sci. 6 No. 1S pp. 105S-109S (2012) Applied Mathematics & Information Sciences An International Journal @ 2012 NSP Natural Sciences Publishing Cor. Two Algorithms of Image Segmentation

More information

Restoring Ink Bleed-Through Degraded Document Images Using a Recursive Unsupervised Classification Technique

Restoring Ink Bleed-Through Degraded Document Images Using a Recursive Unsupervised Classification Technique Restoring Ink Bleed-Through Degraded Document Images Using a Recursive Unsupervised Classification Technique Drira Fadoua, Frank Le Bourgeois, and Hubert Emptoz LIRIS, INSA de LYON, Bâtiment Jules Verne,

More information

Statistic Metrics for Evaluation of Binary Classifiers without Ground-Truth

Statistic Metrics for Evaluation of Binary Classifiers without Ground-Truth Statistic Metrics for Evaluation of Binary Classifiers without Ground-Truth Maksym Fedorchuk, Bart Lamiroy To cite this version: Maksym Fedorchuk, Bart Lamiroy. Statistic Metrics for Evaluation of Binary

More information

Text Enhancement with Asymmetric Filter for Video OCR. Datong Chen, Kim Shearer and Hervé Bourlard

Text Enhancement with Asymmetric Filter for Video OCR. Datong Chen, Kim Shearer and Hervé Bourlard Text Enhancement with Asymmetric Filter for Video OCR Datong Chen, Kim Shearer and Hervé Bourlard Dalle Molle Institute for Perceptual Artificial Intelligence Rue du Simplon 4 1920 Martigny, Switzerland

More information

Text line Segmentation of Curved Document Images

Text line Segmentation of Curved Document Images RESEARCH ARTICLE S OPEN ACCESS Text line Segmentation of Curved Document Images Anusree.M *, Dhanya.M.Dhanalakshmy ** * (Department of Computer Science, Amrita Vishwa Vidhyapeetham, Coimbatore -641 11)

More information

CHAPTER 2 TEXTURE CLASSIFICATION METHODS GRAY LEVEL CO-OCCURRENCE MATRIX AND TEXTURE UNIT

CHAPTER 2 TEXTURE CLASSIFICATION METHODS GRAY LEVEL CO-OCCURRENCE MATRIX AND TEXTURE UNIT CHAPTER 2 TEXTURE CLASSIFICATION METHODS GRAY LEVEL CO-OCCURRENCE MATRIX AND TEXTURE UNIT 2.1 BRIEF OUTLINE The classification of digital imagery is to extract useful thematic information which is one

More information

Fully Automatic Methodology for Human Action Recognition Incorporating Dynamic Information

Fully Automatic Methodology for Human Action Recognition Incorporating Dynamic Information Fully Automatic Methodology for Human Action Recognition Incorporating Dynamic Information Ana González, Marcos Ortega Hortas, and Manuel G. Penedo University of A Coruña, VARPA group, A Coruña 15071,

More information

SEGMENTATION OF CHARACTERS WITHOUT MODIFIERS FROM A PRINTED BANGLA TEXT

SEGMENTATION OF CHARACTERS WITHOUT MODIFIERS FROM A PRINTED BANGLA TEXT SEGMENTATION OF CHARACTERS WITHOUT MODIFIERS FROM A PRINTED BANGLA TEXT ABSTRACT Rupak Bhattacharyya et al. (Eds) : ACER 2013, pp. 11 24, 2013. CS & IT-CSCP 2013 Fakruddin Ali Ahmed Department of Computer

More information

IJSER. various image modalities for non-destructive testing 1. INTRODUCTION (NDT) applications, such as ultrasonic images in, eddy

IJSER. various image modalities for non-destructive testing 1. INTRODUCTION (NDT) applications, such as ultrasonic images in, eddy International Journal of Scientific & Engineering Research Volume 8, Issue 5, May-2017 101 Multilevel Image Thresholding using OTSU s Algorithm in Image Segmentation Priya M.S Research Scholar, Bharathiar

More information

Overcompressing JPEG images with Evolution Algorithms

Overcompressing JPEG images with Evolution Algorithms Author manuscript, published in "EvoIASP2007, Valencia : Spain (2007)" Overcompressing JPEG images with Evolution Algorithms Jacques Lévy Véhel 1, Franklin Mendivil 2 and Evelyne Lutton 1 1 Inria, Complex

More information

Automatic Recognition and Verification of Handwritten Legal and Courtesy Amounts in English Language Present on Bank Cheques

Automatic Recognition and Verification of Handwritten Legal and Courtesy Amounts in English Language Present on Bank Cheques Automatic Recognition and Verification of Handwritten Legal and Courtesy Amounts in English Language Present on Bank Cheques Ajay K. Talele Department of Electronics Dr..B.A.T.U. Lonere. Sanjay L Nalbalwar

More information

Temperature Calculation of Pellet Rotary Kiln Based on Texture

Temperature Calculation of Pellet Rotary Kiln Based on Texture Intelligent Control and Automation, 2017, 8, 67-74 http://www.scirp.org/journal/ica ISSN Online: 2153-0661 ISSN Print: 2153-0653 Temperature Calculation of Pellet Rotary Kiln Based on Texture Chunli Lin,

More information

Image Restoration and Reconstruction

Image Restoration and Reconstruction Image Restoration and Reconstruction Image restoration Objective process to improve an image Recover an image by using a priori knowledge of degradation phenomenon Exemplified by removal of blur by deblurring

More information

Document recto-verso registration using a dynamic time warping algorithm.

Document recto-verso registration using a dynamic time warping algorithm. Document recto-verso registration using a dynamic time warping algorithm. Vincent Rabeux, Nicholas Journet, Jean-Philippe Domenger To cite this version: Vincent Rabeux, Nicholas Journet, Jean-Philippe

More information

Automatic thresholding for defect detection

Automatic thresholding for defect detection Pattern Recognition Letters xxx (2006) xxx xxx www.elsevier.com/locate/patrec Automatic thresholding for defect detection Hui-Fuang Ng * Department of Computer Science and Information Engineering, Asia

More information

OCR For Handwritten Marathi Script

OCR For Handwritten Marathi Script International Journal of Scientific & Engineering Research Volume 3, Issue 8, August-2012 1 OCR For Handwritten Marathi Script Mrs.Vinaya. S. Tapkir 1, Mrs.Sushma.D.Shelke 2 1 Maharashtra Academy Of Engineering,

More information

Image Thresholding using Ultrafuzziness Optimization Based on Type II Fuzzy Sets

Image Thresholding using Ultrafuzziness Optimization Based on Type II Fuzzy Sets Image Thresholding using Ultrafuzziness Optimization Based on Type II Fuzzy Sets gus Zainal rifin, idila Fitri Heddyanna, Hudan Studiawan Vision and Image Processing Laboratory, Department of Informatics,

More information

Comparative Study on Thresholding

Comparative Study on Thresholding omparative Study on Thresholding omparative Study on Thresholding K..Singh #, Lalit Mohan Satapathy *, Bibhudatta Dash *, S.K.Routray * # Department of ET, Orissa Engineering ollege Bhubaneswar, erunalraj@rediffmail.com

More information

Bridging the Gap Between Local and Global Approaches for 3D Object Recognition. Isma Hadji G. N. DeSouza

Bridging the Gap Between Local and Global Approaches for 3D Object Recognition. Isma Hadji G. N. DeSouza Bridging the Gap Between Local and Global Approaches for 3D Object Recognition Isma Hadji G. N. DeSouza Outline Introduction Motivation Proposed Methods: 1. LEFT keypoint Detector 2. LGS Feature Descriptor

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

I. INTRODUCTION. Image Acquisition. Denoising in Wavelet Domain. Enhancement. Binarization. Thinning. Feature Extraction. Matching

I. INTRODUCTION. Image Acquisition. Denoising in Wavelet Domain. Enhancement. Binarization. Thinning. Feature Extraction. Matching A Comparative Analysis on Fingerprint Binarization Techniques K Sasirekha Department of Computer Science Periyar University Salem, Tamilnadu Ksasirekha7@gmail.com K Thangavel Department of Computer Science

More information

Layout Segmentation of Scanned Newspaper Documents

Layout Segmentation of Scanned Newspaper Documents , pp-05-10 Layout Segmentation of Scanned Newspaper Documents A.Bandyopadhyay, A. Ganguly and U.Pal CVPR Unit, Indian Statistical Institute 203 B T Road, Kolkata, India. Abstract: Layout segmentation algorithms

More information

Performance Evaluation of Basic Segmented Algorithms for Brain Tumor Detection

Performance Evaluation of Basic Segmented Algorithms for Brain Tumor Detection IOSR Journal of Electronics and Communication Engineering (IOSR-JECE) e-issn: 2278-2834,p- ISSN: 2278-8735. Volume 5, Issue 6 (Mar. - Apr. 203), PP 08-3 Performance Evaluation of Basic Segmented Algorithms

More information

A QR code identification technology in package auto-sorting system

A QR code identification technology in package auto-sorting system Modern Physics Letters B Vol. 31, Nos. 19 21 (2017) 1740035 (5 pages) c World Scientific Publishing Company DOI: 10.1142/S0217984917400358 A QR code identification technology in package auto-sorting system

More information

Schedule for Rest of Semester

Schedule for Rest of Semester Schedule for Rest of Semester Date Lecture Topic 11/20 24 Texture 11/27 25 Review of Statistics & Linear Algebra, Eigenvectors 11/29 26 Eigenvector expansions, Pattern Recognition 12/4 27 Cameras & calibration

More information