Thai Text Localization in Natural Scene Images using Convolutional Neural Network

Size: px
Start display at page:

Download "Thai Text Localization in Natural Scene Images using Convolutional Neural Network"

Transcription

1 Thai Text Localization in Natural Scene Images using Convolutional Neural Network Thananop Kobchaisawat * and Thanarat H. Chalidabhongse * Department of Computer Engineering, Chulalongkorn University, Bangkok, Thailand Thananop.Ko@student.chula.ac.th Tel: Department of Computer Engineering, Chulalongkorn University, Bangkok, Thailand Thanarat.C@chula.ac.th Tel: Abstract Text detection in natural scene images is a challenging problem due to many variations and uncontrollable factors comparing to text detection on scanned document. Unlike the existing Thai text detection methods which focus on using connected component analysis combining with other rulebased techniques to localize text, our proposed method is based on a well-known automatic feature extractor neural networks called Convolutional Neural Networks (CNN). The CNN is first trained with both English and Thai text datasets. A multi-scaled text confidence maps are constructed in order to cope with the text size variations. Some post-processing and Thai text analysis are also employed to acquire text locations in the image. Base on our experimental results, the proposed method can detect English and Thai text from natural scene images with a promising accuracy comparing to the state-of-the-art method. I. INTRODUCTION Text information in images can be used for many applications, such as automatic language translation, scene text understanding and assistive text reading for visually impaired people. Due to a wide range of applications, this problem has received significant attention from many researchers. However, locating text in natural scene images is unlike locating text from scanned documents. In natural scene images, there are many unpredictable factors, such as text size, style, numerous ranges of backgrounds and the variation of lighting condition. Many methods have been proposed and reported the promising results on English text dataset. Nevertheless, the existing methods do not work well enough on Thai text due to some specific characteristics of the language. Scene text localization techniques can be roughly categorized into two groups which are connected-componentbased and region-based [1]. Connected-component-based methods [2] [4] use prior knowledge of text characteristics such as color, stroke-width, geometric properties combine with post-processing and some heuristics to prune non-text area. Region-based methods typically use sliding window to find text areas in an image. Features are extracted from each window and passed to a classifier to classify text regions. Due to the popularity of machine learning, many region-based text localization methods use machine learning algorithms such as Adaboost [5][6], Support Vector Machine (SVM) [7][8] and Neural Network [9] as classifiers. In order to use these machine learning algorithms, feature extractors are required. Many well-crafted computer vision feature extractors, such as Histogram of Oriented Gradient (HOG), Local Binary Pattern (LBP), and Discrete Wavelet Transform (DWT) are used together with handcrafted features to build up feature extractor. Then, post-processing techniques are applied to acquire text location. In this paper, instead of finding a proper feature extractor for Thai text, we employ Convolutional Neural Network (CNN) which is a kind of learnable feature extractor neural network. CNN has enjoyed many great successes in related fields such as license plate localization, face detection, handwritten digit classification, and character recognition. We are able to create CNN text detector combined with postprocessing techniques to acquire Thai texts location in a natural scene image with promising results. The rest of this paper is organized as follows: In Section II, we present a survey of region-based text localization methods and Thai text localization in natural scenes. Our CNN text detector details are provided in Section III. In Section IV, we present the proposed post-processing techniques, which are designed to improve Thai text localization results. In Section V, we present the experimental results on test datasets compared to other multi-language text localization methods. Finally, we conclude our work in Section VI. II. RELATED WORK Many region-based text localization in natural scene images methods have been proposed. J. Lee et al. [10] presented text detection using six features, which were extracted from multiscaled input image and classified by Modest Adaboost text detector to obtain result map. Then, the post-processing techniques were applied to construct output text region. A text detector proposed by A. Coates et al. [11] used K APSIPA APSIPA 2014

2 mean clustering to learn feature extractor from the preprocessed ICDAR 2003 dataset. Like CNN, this method builds an unsupervised feature extractor from training data. However, there were no details about their text localization post-processing method described in the paper. T.Wang et al. [12] presented an end-to-end scene text recognition including scene text localization using CNN. Their first layer feature extractor of CNN text detector was derived from the method proposed in [11]. A post-processing technique was applied on multi-scaled outputs from CNN text detector to acquire text locations. However, their CNN text detector and the post-processing technique do not work well with Thai text due to some special characteristics of Thai language. For Thai text detection in natural scene images, W. Jirattitichareon and T.H. Chalidabhongse [13] proposed a method to detect Thai text from low-quality signs using edge features, connected component labeling and segmenting with color model using Gaussian Mixture Model (GMM). K.Woraratpunya et al. [14] also introduced Thai text detection from natural scene images using fast boundary clustering and modified connected component analysis with heuristic rules. However, these two methods proposed text analyzing techniques based on some assumptions of specific characteristics of sign images which might not work well on complex scene images containing lots of higher variations of text images. Figure 1. Example of text and non-text patches from ICDAR 2003, ICDAR 2011, Char74k, and SVT datasets. In order to make the CNN detect Thai text more accurately, we needed to train the text detector with Thai text dataset. However, to the best of our knowledge, there is no standard Thai text dataset. Thus, we synthesize images of Thai text dataset from 500 fonts with random sizes, styles, and apply filters to make dataset more realistic. We randomly created 40,000 Thai text patches as shown in Fig. 2. In this work, we use 80,000 text patches (40,000 for English and 40,000 for Thai) and 80,000 non-text patches. III. TEXT DETECTOR In order to localize text in natural scene images using CNN text detector, we had to train our text detector by well-known text dataset combined with synthetic Thai text dataset. Then, the trained text detector was applied on multi-scaled input image to acquire text confidence maps for a post-processing stage. A. Text Detector Learning Architecture 1) Dataset Acquisition and Pre-processing In order to train our text detector, we used 4 well-known text datasets. Those were ICDAR 2003 [15], ICDAR 2011 [16], Char74k [17] and SVT (Street View Text) [18]. These dataset images were converted into grayscale to overcome scene text color variations problem. A 32x32 sliding window was applied on each image to gather text and non-text patches. Each sliding window, that contained text area with at least 80% compared to the provided ground-truth, was counted as a text patch. The sample of acquired dataset is shown in Fig. 1 Figure 2. Example of generated Thai text dataset. Next, a local brightness and contrast normalization was applied to each patch to correct non-uniform illumination and contrast differences, which is defined as,,,, where,,,,, and, represent the output patch, the input patch, the estimated local mean and the estimated local standard deviation of input patch respectively. 2) Text Detector Training The proposed CNN text detector has 5 layers which are 2 convolution layers, 2 average pooling layers, and a fullyconnected layer. The output layer consists of 2 nodes which is text and non-text. The overall text detector model and structure is shown in Fig.3 and Table.1 respectively. (1)

3 Text Input Patch Convolution Layer Non- Text Fully Connected Subsampling Layer Layer Figure 3. CNN Text Detector TABLE I TEXT DETECTOR STRUCTURE Layer Type Input Kernel Feature map / Output Size Size Hidden Layer Size 1 C 32x32 5x x28 2 P 28x28 2x x14 3 C 14x14 3x x12 4 P 12x12 2x x6 5 F 6x6x * C Convolution Layer, P Average Pooling Layer, F Fully-Connected Layer Our text detector was then trained using back-propagation algorithm and with the variance normalized version of hyperbolic tangent sigmoid defined in (2) as a non-linearity function and Mean Squared Error (MSE) as an error function tanh (2) where and represent the output and input of variance normalized version of hyperbolic tangent sigmoid function. B. Text Confidence Map In this part, we used our trained text detector as described above to estimate text locations in the given input image. First, the multi-scaled input image was built. Then, each scaled image was passed to our trained text detector to obtain multi-scaled text confidence maps that were used in the postprocessing stage. 1) Image Pre-Processing An input image was converted into grayscale to deal with text color variations. In order to make our text detector with fixed 32x32 pixel input to detect text in various sizes, an image pyramid was constructed ranging from 10% to 150% of the original input image size with 10% increasing step. 2) Multi-Scaled Text Detector The 32x32 pixel sliding window was performed on each scaled image. A local brightness and contrast normalization was also applied on each window before passing to the trained CNN text detector to obtain multi-scaled text confidence maps as shown in Fig.4. Figure 4. Original input image. Text confidence map at scale 1.5, 1.2, 1.1,0.9,0.8,0.7 (From left to right). IV. TEXT CONFIDENCE MAP POST-PROCESSING In this section, the text confidence maps from Section III were post-processed to obtain final text locations. First, the estimated text bounding boxes were acquired from each text confidence maps. Then, all scaled estimated text bounding boxes were selected based on NMS (Non Maximum Suppression) to suppress overlapping bounding boxes with low scores. Finally, we performed Thai Text analysis on each candidate text boxes to acquire final text bounding boxes. A. Estimated Text Bounding Box A similar technique used in [12] was employed to produce line response. For each scaled text confidence map, we calculated line response by applying line level sliding window rules as defined below,, 0 where denoted line sliding window width. For each row response 0, we constructed line-level bounding box with a proper size to image scale, which made possibility of overlapping bounding box. When a bounding box from all scales was acquired, a NMS was applied to suppress overlapping bounding boxes with low scores and obtain the candidates of text bounding boxes as an output. B. Thai Text Characteristic and Analysis 1) Thai Text Characteristics Thai Text consists of consonants, vowels, tone marks, and special characters as shown in Table II. (3)

4 Unlike English text which all alphabets and vowels are written on a single line, ก ข one ฃ ค line ฅ ฆ of ง จ Thai ฉ ช text ซ ฌ is ญ divided ฎ ฏ ฐ ฑ into ฒ 4 levels as shown in Fig.5. Alphabets and some vowels are Consonants ณ ด ต ถ ท ธ น บ ป ผ ฝ พ ฟ ภ ม ย ร ล written in main level while the others might be written above or below levels. These special ว ศ ษ characteristics ส ห ฬ อ ฮ make the existing state-of-the-art English text detection methods not to work well with Thai text. Vowels Figure 5. Unlike the English text, the characters in Thai text distributive lie over the 4 levels. 2) Thai Text Analysis TABLE 2 THAI TEXT CLASSIFICATION ะาๅำ เโใไ Tone Marks Special Characters ๆฯ In order to refine each candidate text bounding boxes to be more accurate for Thai Text, we must know estimate text line location. We employed Canny edge detector [19] and applied connected component analysis on each candidate text bounding box as shown in Fig Figure 7. Estimated text line layout where a, b and c represent the estimated upper, center and lower text line. Focusing on the center components, we can get the estimated character height from the estimated upper and lower line location. In Thai text, vowels and tone marks are usually written above and below main line characters with height not over than 50% of character height. From this assumptions, we padded candidate text bounding box above and below by 50% of character height from the estimated upper and lower line locations. On the new pad bounding box, we performed same process as above to acquire the estimated center of text line, the upper and lower line locations. We considered the components above upper line and below lower line as upper and lower components respectively. a b c Upper Center Lower Figure 8. Padded candidate text bounding box. Text layout analysis result. (c) Figure 6. Candidate text bounding box. Canny edge detector result. (c) Connected Component Analysis result. We computed the centroid of each connected component and found the mean of y-position as estimated center of text line. The connected components which lied in ± 30% interval of estimated center line were considered as center line components. Then, estimated upper and lower line location were calculated from the center line components. From this information, we were able to acquire text line layout as shown in Fig.7 to perform components analysis. For each upper and lower components, we found the center line component as shown in Fig.8b. It had the least distance among the other center components. Normally, Thai text places the upper and lower components above or below center line character between 0 to 45 degrees from center line character centroid as shown in Fig.9. This hypothesis helped us consider that should addition components from padded text bounding box be included in final text bounding box.

5 0 45 TABLE 4 TEXT LOCALIZATION METHODS PERFORMANCE EVALUATION ON ICDAR2003 DATASET Method Dataset English [ICDAR2003] Precision Recall F-Measure 1 st ICDAR [15] Epshtein [2] Proposed method B.Bai [20] Y.Pan [3] Figure 9. Example of proper align upper component. After we calculated all upper and lower components hypothesis. For the final text bounding box, we padded the upper and lower of bounding box only if >70% of upper and lower components were aligned in proper positions. Then, we built the minimal bounding box, which included all proper components as the final text bounding box as shown in Fig.10. Figure 10. The estimated text bounding box. The final refined text bounding box. V. EXPERIMENTAL RESULT In order to evaluate our proposed method, we had conducted the experiment using Thai-English text and English text only dataset on different methods. Our Thai-English text dataset consists of 200 images in 640x480 pixel and English dataset is from ICDAR 2003 and ICDAR 2011 standard test dataset. We evaluated performance using 3 ICDAR standard text localization evaluation criterions [15] which are precision, recall and f-measure between our proposed method and other multi-language text localization methods. The result on each dataset is shown in Table 3-5. TABLE 3 TEXT LOCALIZATION METHODS PERFORMANCE EVALUATION ON THAI-ENGLISH DATASET Method Dataset Thai-English Precision Recall F-Measure Epshtein [2] T.Wang [12] Proposed method TABLE 5 TEXT LOCALIZATION METHODS PERFORMANCE EVALUATION ON ICDAR2011 DATASET Method Dataset English [ICDAR2011] Precision Recall F-Measure C.Yi [21] Proposed method Epshtein [1] L. Neumann [1] In Fig.11, we compared the results of each text localization method on Thai-English dataset. From the result, our method detected and localized Thai text more accurately. Our proposed method results in Fig.11 (e) showed that the vowels and tone marks in each test images were included in text bonding box compared to other multi-language text localization methods, which can detect center line characters but miss some vowels and tones marks. Fig.12 are some sample outputs from our proposed method. We can see that our method can detect both Thai and English natural scene text in variations of text styles, sizes, and colors, and even with a small effect of perspective distortion. However, our method failed to localize text in some difficult cases, such as texts that are significantly distorted due to perspective projection, text on non-linear planar, too small text, text on complex background and lightning condition. We show failed results in Fig.13 VI. CONCLUSIONS In this paper, we present a method to localize Thai text in natural scene images. Our system uses the learned features neural network (CNN) as text detector combined with the post-processing techniques with Thai text characteristic analysis. This combination improved Thai text localization result in term of less missed vowels and tones compared to other multi-language text localization methods. From the experimental result, based on standard evaluation method, our system shows good result on mixed Thai and English test images.

6 (c) (d) (e) Figure 11. Input image and interested text area. Result from T.Wang s method. (c) Result from Epshtein s method. (d) Result from L.Neumann s method ( (e) Our proposed method result. REFERENCES Figure 12. Example of correct results. [1] L. Neumann and J. Matas, Scene Text Localization and Recognition with Oriented Stroke Detection, in IEEE International Conference on Computer Vision, (ICCV2013), 2013, pp [2] B. Epshtein, E. Ofek, and Y. Wexler, Detecting text in natural scenes with stroke width transform, in IEEE Conference on Computer Vision and Pattern Recognition, (CVPR2012), 2010, pp [3] Y.-F. Pan, X. Hou, and C.-L. Liu, A Hybrid Approach to Detect and Localize Texts in Natural Scene Images, IEEE Trans. Image Process., vol. 20, no. 3, pp , [4] L. Neumann and J. Matas, A Method for Text Localization and Recognition in Real-world Images, in Computer Vision ACCV 2010, 2010, pp [5] X. Chen and A. L. Yuille, Detecting and reading text in natural scenes, in IEEE Conference on Computer Vision and Pattern Recognition, (CVPR2004), 2004, vol. 2, pp. II 366 II 373 Vol.2. [6] Y.-F. Pan, X. Hou, and C.-L. Liu, A Robust System to Detect and Localize Texts in Natural Scene Images, in 8th IAPR International Workshop on Document Analysis Systems, (DAS2008), 2008, pp [7] D. Chen, H. Bourlard, and J. Thiran, Text identification in complex background using SVM, in IEEE Conference on Computer Vision and Pattern Recognition, (CVPR 2001), 2001, vol. 2, pp. II 621 II 626 vol.2. [8] X. Li, W. Wang, S. Jiang, Q. Huang, and W. Gao, Fast and effective text detection, in 15th IEEE International Conference on

7 Image Processing, (ICIP2008), 2008, pp [9] S. M. Hanif and L. Prevost, Text Detection and Localization in Complex Scene Images using Constrained AdaBoost Algorithm, in 10th International Conference on Document Analysis and Recognition, (ICDAR2009), 2009, pp [10] J.-J. Lee, P.-H. Lee, S.-W. Lee, A. Yuille, and C. Koch, AdaBoost for Text Detection in Natural Scene, in 11th International Conference on Document Analysis and Recognition, (ICDAR2011), 2011, pp [11] A. Coates, B. Carpenter, C. Case, S. Satheesh, B. Suresh, W. Tao, and A. Y. Ng, Text Detection and Character Recognition in Scene Images with Unsupervised Feature Learning, in 11th International Conference on Document Analysis and Recognition, (ICDAR2011), 2011, pp [12] T. Wang, D. J. Wu, A. Coates, and A. Y. Ng, End-to-end text recognition with convolutional neural networks, in 21st International Conference on Pattern Recognition,2012 (ICPR2012), 2012, pp [13] W. Jirattitichareon and T. H. Chalidabhongse, Automatic Detection and Segmentation of Text in Low Quality Thai Sign Images, in IEEE Asia-Pacific Conference on Circuits and Systems, (APCCAS2006), 2006, pp [14] K. Woraratpanya, P. Boonchukusol, Y. Kuroki, and Y. Kato, Improved Thai text detection from natural scenes, in International Conference on Information Technology and Electrical Engineering, (ICITEE2013), 2013, pp [15] S. M. Lucas, A. Panaretos, L. Sosa, A. Tang, S. Wong, and R. Young, ICDAR 2003 robust reading competitions, in 7th International Conference on Document Analysis and Recognition, (ICDAR2003), 2003, vol. 1, no. Icdar, pp [16] A. Shahab, F. Shafait, and A. Dengel, ICDAR 2011 Robust Reading Competition Challenge 2: Reading Text in Scene Images, in 11th International Conference on Document Analysis and Recognition, (ICDAR2011), 2011, pp [17] T. E. de Campos, B. R. Babu, and M. Varma, Character recognition in natural images, in Proceedings of the International Conference on Computer Vision Theory and Applications, Lisbon, Portugal, [18] K. Wang, B. Babenko, and S. Belongie, End-to-end scene text recognition, in 13th International Conference on Computer Vision, (ICCV2011), 2011, pp [19] J. Canny, A Computational Approach to Edge Detection, IEEE Trans. Pattern Anal. Mach. Intell., vol. PAMI-8, no. 6, pp , [20] B. Bai, F. Yin, and C. L. Liu, Scene Text Localization Using Gradient Local Correlation, in 12th International Conference on Document Analysis and Recognition, (ICDAR2013), 2013, pp [21] C. Yi and Y. Tian, Text String Detection From Natural Scenes by Structure-Based Partition and Grouping, IEEE Trans. Image Process., vol. 20, no. 9, pp , Figure 13. Example of incorrect results.

Broken Characters Identification for Thai Character Recognition Systems

Broken Characters Identification for Thai Character Recognition Systems Broken Characters Identification for Thai Character Recognition Systems NUCHAREE PREMCHAISWADI*, WICHIAN PREMCHAISWADI* UBOLRAT PACHIYANUKUL**, SEINOSUKE NARITA*** *Faculty of Information Technology, Dhurakijpundit

More information

The Clustering Technique for Thai Handwritten Recognition

The Clustering Technique for Thai Handwritten Recognition The Clustering Technique for Thai Handwritten Recognition Ithipan Methasate, Sutat Sae-tang Information Research and Development Division National Electronics and Computer Technology Center National Science

More information

Detecting and Recognizing Text in Natural Images using Convolutional Networks

Detecting and Recognizing Text in Natural Images using Convolutional Networks Detecting and Recognizing Text in Natural Images using Convolutional Networks Aditya Srinivas Timmaraju, Vikesh Khanna Stanford University Stanford, CA - 94305 adityast@stanford.edu, vikesh@stanford.edu

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

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

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

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

LEVERAGING SURROUNDING CONTEXT FOR SCENE TEXT DETECTION

LEVERAGING SURROUNDING CONTEXT FOR SCENE TEXT DETECTION LEVERAGING SURROUNDING CONTEXT FOR SCENE TEXT DETECTION Yao Li 1, Chunhua Shen 1, Wenjing Jia 2, Anton van den Hengel 1 1 The University of Adelaide, Australia 2 University of Technology, Sydney, Australia

More information

LETTER Learning Co-occurrence of Local Spatial Strokes for Robust Character Recognition

LETTER Learning Co-occurrence of Local Spatial Strokes for Robust Character Recognition IEICE TRANS. INF. & SYST., VOL.E97 D, NO.7 JULY 2014 1937 LETTER Learning Co-occurrence of Local Spatial Strokes for Robust Character Recognition Song GAO, Student Member, Chunheng WANG a), Member, Baihua

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

Multi-script Text Extraction from Natural Scenes

Multi-script Text Extraction from Natural Scenes Multi-script Text Extraction from Natural Scenes Lluís Gómez and Dimosthenis Karatzas Computer Vision Center Universitat Autònoma de Barcelona Email: {lgomez,dimos}@cvc.uab.es Abstract Scene text extraction

More information

Available online at ScienceDirect. Procedia Computer Science 96 (2016 )

Available online at   ScienceDirect. Procedia Computer Science 96 (2016 ) Available online at www.sciencedirect.com ScienceDirect Procedia Computer Science 96 (2016 ) 1409 1417 20th International Conference on Knowledge Based and Intelligent Information and Engineering Systems,

More information

Hand Motion Analysis for Thai Alphabet Recognition using HMM

Hand Motion Analysis for Thai Alphabet Recognition using HMM Hand Motion Analysis for Thai Alphabet Recognition using HMM Kittasil Silanon and Nikom Suvonvorn Abstract In this paper, we propose a system for Thai alphabet recognition from hand movement trajectory,

More information

Text Detection and Character Recognition in Scene Images with Unsupervised Feature Learning

Text Detection and Character Recognition in Scene Images with Unsupervised Feature Learning Text Detection and Character Recognition in Scene Images with Unsupervised Feature Learning Adam Coates, Blake Carpenter, Carl Case, Sanjeev Satheesh, Bipin Suresh, Tao Wang, Andrew Y. Ng Computer Science

More information

Scene Text Recognition in Mobile Applications by Character Descriptor and Structure Configuration

Scene Text Recognition in Mobile Applications by Character Descriptor and Structure Configuration 1 Scene Text Recognition in Mobile Applications by Character Descriptor and Structure Configuration Chucai Yi, Student Member, IEEE, and Yingli Tian, Senior Member, IEEE Abstract Text characters and strings

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

Racing Bib Number Recognition

Racing Bib Number Recognition BEN-AMI, BASHA, AVIDAN: RACING BIB NUMBER RECOGNITION 1 Racing Bib Number Recognition Idan Ben-Ami idan.benami@gmail.com Tali Basha talib@eng.tau.ac.il Shai Avidan avidan@eng.tau.ac.il School of Electrical

More information

A Hierarchical Visual Saliency Model for Character Detection in Natural Scenes

A Hierarchical Visual Saliency Model for Character Detection in Natural Scenes A Hierarchical Visual Saliency Model for Character Detection in Natural Scenes Renwu Gao 1, Faisal Shafait 2, Seiichi Uchida 3, and Yaokai Feng 3 1 Information Sciene and Electrical Engineering, Kyushu

More information

Text Detection in Multi-Oriented Natural Scene Images

Text Detection in Multi-Oriented Natural Scene Images Text Detection in Multi-Oriented Natural Scene Images M. Fouzia 1, C. Shoba Bindu 2 1 P.G. Student, Department of CSE, JNTU College of Engineering, Anantapur, Andhra Pradesh, India 2 Associate Professor,

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

Text Detection from Natural Image using MSER and BOW

Text Detection from Natural Image using MSER and BOW International Journal of Emerging Engineering Research and Technology Volume 3, Issue 11, November 2015, PP 152-156 ISSN 2349-4395 (Print) & ISSN 2349-4409 (Online) Text Detection from Natural Image using

More information

arxiv: v1 [cs.cv] 23 Apr 2016

arxiv: v1 [cs.cv] 23 Apr 2016 Text Flow: A Unified Text Detection System in Natural Scene Images Shangxuan Tian1, Yifeng Pan2, Chang Huang2, Shijian Lu3, Kai Yu2, and Chew Lim Tan1 arxiv:1604.06877v1 [cs.cv] 23 Apr 2016 1 School of

More information

Detection of Text with Connected Component Clustering

Detection of Text with Connected Component Clustering Detection of Text with Connected Component Clustering B.Nishanthi 1, S. Shahul Hammed 2 PG Scholar, Dept. Computer Science and Engineering, Karpagam University, Coimbatore, Tamil Nadu, India 1 Assistant

More information

A Fast Caption Detection Method for Low Quality Video Images

A Fast Caption Detection Method for Low Quality Video Images 2012 10th IAPR International Workshop on Document Analysis Systems A Fast Caption Detection Method for Low Quality Video Images Tianyi Gui, Jun Sun, Satoshi Naoi Fujitsu Research & Development Center CO.,

More information

Scene text extraction based on edges and support vector regression

Scene text extraction based on edges and support vector regression IJDAR (2015) 18:125 135 DOI 10.1007/s10032-015-0237-z SPECIAL ISSUE PAPER Scene text extraction based on edges and support vector regression Shijian Lu Tao Chen Shangxuan Tian Joo-Hwee Lim Chew-Lim Tan

More information

Scene text recognition: no country for old men?

Scene text recognition: no country for old men? Scene text recognition: no country for old men? Lluís Gómez and Dimosthenis Karatzas Computer Vision Center Universitat Autònoma de Barcelona Email: {lgomez,dimos}@cvc.uab.es Abstract. It is a generally

More information

A Skeleton Based Descriptor for Detecting Text in Real Scene Images

A Skeleton Based Descriptor for Detecting Text in Real Scene Images A Skeleton Based Descriptor for Detecting Text in Real Scene Images Mehdi Felhi, Nicolas Bonnier, Salvatore Tabbone To cite this version: Mehdi Felhi, Nicolas Bonnier, Salvatore Tabbone. A Skeleton Based

More information

Towards Visual Words to Words

Towards Visual Words to Words Towards Visual Words to Words Text Detection with a General Bag of Words Representation Rakesh Mehta Dept. of Signal Processing, Tampere Univ. of Technology in Tampere Ondřej Chum, Jiří Matas Centre for

More information

12/12 A Chinese Words Detection Method in Camera Based Images Qingmin Chen, Yi Zhou, Kai Chen, Li Song, Xiaokang Yang Institute of Image Communication

12/12 A Chinese Words Detection Method in Camera Based Images Qingmin Chen, Yi Zhou, Kai Chen, Li Song, Xiaokang Yang Institute of Image Communication A Chinese Words Detection Method in Camera Based Images Qingmin Chen, Yi Zhou, Kai Chen, Li Song, Xiaokang Yang Institute of Image Communication and Information Processing, Shanghai Key Laboratory Shanghai

More information

Image Text Extraction and Recognition using Hybrid Approach of Region Based and Connected Component Methods

Image Text Extraction and Recognition using Hybrid Approach of Region Based and Connected Component Methods Image Text Extraction and Recognition using Hybrid Approach of Region Based and Connected Component Methods Ms. N. Geetha 1 Assistant Professor Department of Computer Applications Vellalar College for

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

WITH the increasing use of digital image capturing

WITH the increasing use of digital image capturing 800 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 20, NO. 3, MARCH 2011 A Hybrid Approach to Detect and Localize Texts in Natural Scene Images Yi-Feng Pan, Xinwen Hou, and Cheng-Lin Liu, Senior Member, IEEE

More information

Automatically Algorithm for Physician s Handwritten Segmentation on Prescription

Automatically Algorithm for Physician s Handwritten Segmentation on Prescription Automatically Algorithm for Physician s Handwritten Segmentation on Prescription Narumol Chumuang 1 and Mahasak Ketcham 2 Department of Information Technology, Faculty of Information Technology, King Mongkut's

More information

Aggregating Local Context for Accurate Scene Text Detection

Aggregating Local Context for Accurate Scene Text Detection Aggregating Local Context for Accurate Scene Text Detection Dafang He 1, Xiao Yang 2, Wenyi Huang, 1, Zihan Zhou 1, Daniel Kifer 2, and C.Lee Giles 1 1 Information Science and Technology, Penn State University

More information

A Hybrid Approach To Detect And Recognize Text In Images

A Hybrid Approach To Detect And Recognize Text In Images IOSR Journal of Engineering (IOSRJEN) ISSN (e): 2250-3021, ISSN (p): 2278-8719 Vol. 04, Issue 07 (July. 2014), V4 PP 19-20 www.iosrjen.org A Hybrid Approach To Detect And Recognize Text In Images 1 Miss.Poonam

More information

ค ม อการใช งาน ET-BASE GSM SIM900

ค ม อการใช งาน ET-BASE GSM SIM900 ค ม อการใช งาน ET-BASE GSM SIM900 Electronics Technology Team บร ษ ท อ ท ท จำก ด ETT CO., LTD. http://www.etteam.com CONTENTS 1. Specifications of Board ET-BASE GSM SIM900 3 2. Initial Specifications of

More information

SCENE TEXT RECOGNITION IN MULTIPLE FRAMES BASED ON TEXT TRACKING

SCENE TEXT RECOGNITION IN MULTIPLE FRAMES BASED ON TEXT TRACKING SCENE TEXT RECOGNITION IN MULTIPLE FRAMES BASED ON TEXT TRACKING Xuejian Rong 1, Chucai Yi 2, Xiaodong Yang 1 and Yingli Tian 1,2 1 The City College, 2 The Graduate Center, City University of New York

More information

INTELLIGENT transportation systems have a significant

INTELLIGENT transportation systems have a significant INTL JOURNAL OF ELECTRONICS AND TELECOMMUNICATIONS, 205, VOL. 6, NO. 4, PP. 35 356 Manuscript received October 4, 205; revised November, 205. DOI: 0.55/eletel-205-0046 Efficient Two-Step Approach for Automatic

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

Face Detection Using Convolutional Neural Networks and Gabor Filters

Face Detection Using Convolutional Neural Networks and Gabor Filters Face Detection Using Convolutional Neural Networks and Gabor Filters Bogdan Kwolek Rzeszów University of Technology W. Pola 2, 35-959 Rzeszów, Poland bkwolek@prz.rzeszow.pl Abstract. This paper proposes

More information

Gradient Difference Based Approach for Text Localization in Compressed Domain

Gradient Difference Based Approach for Text Localization in Compressed Domain Proceedings of International Conference on Emerging Research in Computing, Information, Communication and Applications (ERCICA-14) Gradient Difference Based Approach for Text Localization in Compressed

More information

ZONE: LOOX LED No. Art No. Description Unit Unit Rate PANEL 1: COMPLETE 12V / 24V / 350MA SOLUTIONS

ZONE: LOOX LED No. Art No. Description Unit Unit Rate PANEL 1: COMPLETE 12V / 24V / 350MA SOLUTIONS ZONE: LOOX LED PANEL 1: COMPLETE 12V / 24V / 350MA SOLUTIONS A 833.77.903 LED POWER SUPPLY 24V/30W WITH EU-PLUG 1 PC 1100 B 833.77.090 LED3021 SURFACE MOUNTED LIGHT 24V/5.5W CW 4000K BAR SHAPED IP44 265MM

More information

International Journal of Electrical, Electronics ISSN No. (Online): and Computer Engineering 3(2): 85-90(2014)

International Journal of Electrical, Electronics ISSN No. (Online): and Computer Engineering 3(2): 85-90(2014) I J E E E C International Journal of Electrical, Electronics ISSN No. (Online): 2277-2626 Computer Engineering 3(2): 85-90(2014) Robust Approach to Recognize Localize Text from Natural Scene Images Khushbu

More information

Image Retrieval System for Composite Images using Directional Chain Codes

Image Retrieval System for Composite Images using Directional Chain Codes , pp.51-64 http://dx.doi.org/10.14257/ijast.2013.58.05 Image Retrieval System for Composite Images using Directional Chain Codes Akriti Nigam*, Rupesh Yadav* and R. C. Tripathi* *Indian Institute of Information

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

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

Arbitrary-Oriented Scene Text Detection via Rotation Proposals

Arbitrary-Oriented Scene Text Detection via Rotation Proposals 1 Arbitrary-Oriented Scene Text Detection via Rotation Proposals Jianqi Ma, Weiyuan Shao, Hao Ye, Li Wang, Hong Wang, Yingbin Zheng, Xiangyang Xue arxiv:1703.01086v1 [cs.cv] 3 Mar 2017 Abstract This paper

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

Dot Text Detection Based on FAST Points

Dot Text Detection Based on FAST Points Dot Text Detection Based on FAST Points Yuning Du, Haizhou Ai Computer Science & Technology Department Tsinghua University Beijing, China dyn10@mails.tsinghua.edu.cn, ahz@mail.tsinghua.edu.cn Shihong Lao

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

Robust Text Reading in Natural Scene Images

Robust Text Reading in Natural Scene Images Robust Text Reading in Natural Scene Images Tao Wang, David Wu Stanford Computer Science Department 353 Serra Mall, Stanford, CA 94305 twangcat@stanford.edu, dwu4@stanford.edu Abstract In this paper, we

More information

arxiv: v1 [cs.cv] 4 Dec 2017

arxiv: v1 [cs.cv] 4 Dec 2017 Enhanced Characterness for Text Detection in the Wild Aarushi Agrawal 2, Prerana Mukherjee 1, Siddharth Srivastava 1, and Brejesh Lall 1 arxiv:1712.04927v1 [cs.cv] 4 Dec 2017 1 Department of Electrical

More information

Research Article International Journals of Advanced Research in Computer Science and Software Engineering ISSN: X (Volume-7, Issue-7)

Research Article International Journals of Advanced Research in Computer Science and Software Engineering ISSN: X (Volume-7, Issue-7) International Journals of Advanced Research in Computer Science and Software Engineering ISSN: 2277-128X (Volume-7, Issue-7) Research Article July 2017 Technique for Text Region Detection in Image Processing

More information

Part-Based Skew Estimation for Mathematical Expressions

Part-Based Skew Estimation for Mathematical Expressions Soma Shiraishi, Yaokai Feng, and Seiichi Uchida shiraishi@human.ait.kyushu-u.ac.jp {fengyk,uchida}@ait.kyushu-u.ac.jp Abstract We propose a novel method for the skew estimation on text images containing

More information

Handwritten Hindi Numerals Recognition System

Handwritten Hindi Numerals Recognition System CS365 Project Report Handwritten Hindi Numerals Recognition System Submitted by: Akarshan Sarkar Kritika Singh Project Mentor: Prof. Amitabha Mukerjee 1 Abstract In this project, we consider the problem

More information

Object Detection. Part1. Presenter: Dae-Yong

Object Detection. Part1. Presenter: Dae-Yong Object Part1 Presenter: Dae-Yong Contents 1. What is an Object? 2. Traditional Object Detector 3. Deep Learning-based Object Detector What is an Object? Subset of Object Recognition What is an Object?

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

Recognizing Text in the Wild

Recognizing Text in the Wild Bachelor thesis Computer Science Radboud University Recognizing Text in the Wild Author: Twan Cuijpers s4378911 First supervisor/assessor: dr. Twan van Laarhoven T.vanLaarhoven@cs.ru.nl Second assessor:

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

A process for text recognition of generic identification documents over cloud computing

A process for text recognition of generic identification documents over cloud computing 142 Int'l Conf. IP, Comp. Vision, and Pattern Recognition IPCV'16 A process for text recognition of generic identification documents over cloud computing Rodolfo Valiente, Marcelo T. Sadaike, José C. Gutiérrez,

More information

Three-Dimensional Object Detection and Layout Prediction using Clouds of Oriented Gradients

Three-Dimensional Object Detection and Layout Prediction using Clouds of Oriented Gradients ThreeDimensional Object Detection and Layout Prediction using Clouds of Oriented Gradients Authors: Zhile Ren, Erik B. Sudderth Presented by: Shannon Kao, Max Wang October 19, 2016 Introduction Given an

More information

Robust Face Detection Based on Convolutional Neural Networks

Robust Face Detection Based on Convolutional Neural Networks Robust Face Detection Based on Convolutional Neural Networks M. Delakis and C. Garcia Department of Computer Science, University of Crete P.O. Box 2208, 71409 Heraklion, Greece {delakis, cgarcia}@csd.uoc.gr

More information

Efficient indexing for Query By String text retrieval

Efficient indexing for Query By String text retrieval Efficient indexing for Query By String text retrieval Suman K. Ghosh Lluís, Gómez, Dimosthenis Karatzas and Ernest Valveny Computer Vision Center, Dept. Ciències de la Computació Universitat Autònoma de

More information

Deep Tracking: Biologically Inspired Tracking with Deep Convolutional Networks

Deep Tracking: Biologically Inspired Tracking with Deep Convolutional Networks Deep Tracking: Biologically Inspired Tracking with Deep Convolutional Networks Si Chen The George Washington University sichen@gwmail.gwu.edu Meera Hahn Emory University mhahn7@emory.edu Mentor: Afshin

More information

SSD: Single Shot MultiBox Detector. Author: Wei Liu et al. Presenter: Siyu Jiang

SSD: Single Shot MultiBox Detector. Author: Wei Liu et al. Presenter: Siyu Jiang SSD: Single Shot MultiBox Detector Author: Wei Liu et al. Presenter: Siyu Jiang Outline 1. Motivations 2. Contributions 3. Methodology 4. Experiments 5. Conclusions 6. Extensions Motivation Motivation

More information

Real-Time Scene Text Localization and Recognition

Real-Time Scene Text Localization and Recognition Real-Time Scene Text Localization and Recognition Lukáš Neumann Jiří Matas Centre for Machine Perception, Department of Cybernetics Czech Technical University, Prague, Czech Republic neumalu1@cmp.felk.cvut.cz,

More information

Convolution Neural Networks for Chinese Handwriting Recognition

Convolution Neural Networks for Chinese Handwriting Recognition Convolution Neural Networks for Chinese Handwriting Recognition Xu Chen Stanford University 450 Serra Mall, Stanford, CA 94305 xchen91@stanford.edu Abstract Convolutional neural networks have been proven

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

TRANSPARENT OBJECT DETECTION USING REGIONS WITH CONVOLUTIONAL NEURAL NETWORK

TRANSPARENT OBJECT DETECTION USING REGIONS WITH CONVOLUTIONAL NEURAL NETWORK TRANSPARENT OBJECT DETECTION USING REGIONS WITH CONVOLUTIONAL NEURAL NETWORK 1 Po-Jen Lai ( 賴柏任 ), 2 Chiou-Shann Fuh ( 傅楸善 ) 1 Dept. of Electrical Engineering, National Taiwan University, Taiwan 2 Dept.

More information

Scene Text Recognition in Mobile Application using K-Mean Clustering and Support Vector Machine

Scene Text Recognition in Mobile Application using K-Mean Clustering and Support Vector Machine ISSN: 2278 1323 All Rights Reserved 2015 IJARCET 2492 Scene Text Recognition in Mobile Application using K-Mean Clustering and Support Vector Machine Priyanka N Guttedar, Pushpalata S Abstract In natural

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

Face Recognition Using Vector Quantization Histogram and Support Vector Machine Classifier Rong-sheng LI, Fei-fei LEE *, Yan YAN and Qiu CHEN

Face Recognition Using Vector Quantization Histogram and Support Vector Machine Classifier Rong-sheng LI, Fei-fei LEE *, Yan YAN and Qiu CHEN 2016 International Conference on Artificial Intelligence: Techniques and Applications (AITA 2016) ISBN: 978-1-60595-389-2 Face Recognition Using Vector Quantization Histogram and Support Vector Machine

More information

2 OVERVIEW OF RELATED WORK

2 OVERVIEW OF RELATED WORK Utsushi SAKAI Jun OGATA This paper presents a pedestrian detection system based on the fusion of sensors for LIDAR and convolutional neural network based image classification. By using LIDAR our method

More information

Deformable Part Models

Deformable Part Models CS 1674: Intro to Computer Vision Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 9, 2016 Today: Object category detection Window-based approaches: Last time: Viola-Jones

More information

TRAFFIC SIGN RECOGNITION USING A MULTI-TASK CONVOLUTIONAL NEURAL NETWORK

TRAFFIC SIGN RECOGNITION USING A MULTI-TASK CONVOLUTIONAL NEURAL NETWORK TRAFFIC SIGN RECOGNITION USING A MULTI-TASK CONVOLUTIONAL NEURAL NETWORK Dr. S.V. Shinde Arshiya Sayyad Uzma Shaikh Department of IT Department of IT Department of IT Pimpri Chinchwad college of Engineering

More information

Stacked Denoising Autoencoders for Face Pose Normalization

Stacked Denoising Autoencoders for Face Pose Normalization Stacked Denoising Autoencoders for Face Pose Normalization Yoonseop Kang 1, Kang-Tae Lee 2,JihyunEun 2, Sung Eun Park 2 and Seungjin Choi 1 1 Department of Computer Science and Engineering Pohang University

More information

Classification of objects from Video Data (Group 30)

Classification of objects from Video Data (Group 30) Classification of objects from Video Data (Group 30) Sheallika Singh 12665 Vibhuti Mahajan 12792 Aahitagni Mukherjee 12001 M Arvind 12385 1 Motivation Video surveillance has been employed for a long time

More information

MSCS with Distinction in Research Final Report Scene Text Recognition with Convolutional Neural Networks

MSCS with Distinction in Research Final Report Scene Text Recognition with Convolutional Neural Networks MSCS with Distinction in Research Final Report Scene Text Recognition with Convolutional Neural Networks Tao Wang Stanford University, 353 Serra Mall, Stanford, CA 94305 twangcat@cs.stanford.edu Primary

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

Using Adaptive Run Length Smoothing Algorithm for Accurate Text Localization in Images

Using Adaptive Run Length Smoothing Algorithm for Accurate Text Localization in Images Using Adaptive Run Length Smoothing Algorithm for Accurate Text Localization in Images Martin Rais, Norberto A. Goussies, and Marta Mejail Departamento de Computación, Facultad de Ciencias Exactas y Naturales,

More information

A Background Modeling Approach Based on Visual Background Extractor Taotao Liu1, a, Lin Qi2, b and Guichi Liu2, c

A Background Modeling Approach Based on Visual Background Extractor Taotao Liu1, a, Lin Qi2, b and Guichi Liu2, c 4th International Conference on Mechatronics, Materials, Chemistry and Computer Engineering (ICMMCCE 2015) A Background Modeling Approach Based on Visual Background Extractor Taotao Liu1, a, Lin Qi2, b

More information

Extraction and Classification of User Interface Components from an Image

Extraction and Classification of User Interface Components from an Image Volume 118 No. 24 2018 ISSN: 1314-3395 (on-line version) url: http://www.acadpubl.eu/hub/ http://www.acadpubl.eu/hub/ Extraction and Classification of User Interface Components from an Image Saad Hassan

More information

Traffic Sign Localization and Classification Methods: An Overview

Traffic Sign Localization and Classification Methods: An Overview Traffic Sign Localization and Classification Methods: An Overview Ivan Filković University of Zagreb Faculty of Electrical Engineering and Computing Department of Electronics, Microelectronics, Computer

More information

Cursive Handwriting Recognition System Using Feature Extraction and Artificial Neural Network

Cursive Handwriting Recognition System Using Feature Extraction and Artificial Neural Network Cursive Handwriting Recognition System Using Feature Extraction and Artificial Neural Network Utkarsh Dwivedi 1, Pranjal Rajput 2, Manish Kumar Sharma 3 1UG Scholar, Dept. of CSE, GCET, Greater Noida,

More information

Perspective Scene Text Recognition With Feature Compression and Ranking

Perspective Scene Text Recognition With Feature Compression and Ranking Perspective Scene Text Recognition With Feature Compression and Ranking Yu Zhou 1, Shuang Liu 1, Yongzheng Zhang 1, Yipeng Wang 1, and Weiyao Lin 2 1 Institute of Information Engineering, Chinese Academy

More information

Video Inter-frame Forgery Identification Based on Optical Flow Consistency

Video Inter-frame Forgery Identification Based on Optical Flow Consistency Sensors & Transducers 24 by IFSA Publishing, S. L. http://www.sensorsportal.com Video Inter-frame Forgery Identification Based on Optical Flow Consistency Qi Wang, Zhaohong Li, Zhenzhen Zhang, Qinglong

More information

An Efficient Learning Scheme for Extreme Learning Machine and Its Application

An Efficient Learning Scheme for Extreme Learning Machine and Its Application An Efficient Learning Scheme for Extreme Learning Machine and Its Application Kheon-Hee Lee, Miso Jang, Keun Park, Dong-Chul Park, Yong-Mu Jeong and Soo-Young Min Abstract An efficient learning scheme

More information

Human Detection and Tracking for Video Surveillance: A Cognitive Science Approach

Human Detection and Tracking for Video Surveillance: A Cognitive Science Approach Human Detection and Tracking for Video Surveillance: A Cognitive Science Approach Vandit Gajjar gajjar.vandit.381@ldce.ac.in Ayesha Gurnani gurnani.ayesha.52@ldce.ac.in Yash Khandhediya khandhediya.yash.364@ldce.ac.in

More information

Learning-Based Candidate Segmentation Scoring for Real-Time Recognition of Online Overlaid Chinese Handwriting

Learning-Based Candidate Segmentation Scoring for Real-Time Recognition of Online Overlaid Chinese Handwriting 2013 12th International Conference on Document Analysis and Recognition Learning-Based Candidate Segmentation Scoring for Real-Time Recognition of Online Overlaid Chinese Handwriting Yan-Fei Lv 1, Lin-Lin

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

Scene Text Recognition for Augmented Reality. Sagar G V Adviser: Prof. Bharadwaj Amrutur Indian Institute Of Science

Scene Text Recognition for Augmented Reality. Sagar G V Adviser: Prof. Bharadwaj Amrutur Indian Institute Of Science Scene Text Recognition for Augmented Reality Sagar G V Adviser: Prof. Bharadwaj Amrutur Indian Institute Of Science Outline Research area and motivation Finding text in natural scenes Prior art Improving

More information

Text Localization in Real-world Images using Efficiently Pruned Exhaustive Search

Text Localization in Real-world Images using Efficiently Pruned Exhaustive Search Text Localization in Real-world Images using Efficiently Pruned Exhaustive Search Lukáš Neumann Centre for Machine Perception, Dept. of Cybernetics Czech Technical University, Prague, Czech Republic neumalu@cmp.felk.cvut.cz

More information

Proposal for the Thai Script Root Zone LGR

Proposal for the Thai Script Root Zone LGR LGR LGR Version: 2 Date: 2017 05 25 Document version: 6.9b Authors: Contents 1 General Information/ Overview/ Abstract... 3 2 Script for which the LGR is proposed... 3 3 Background on Script and Principal

More information

SAFE: Scale Aware Feature Encoder for Scene Text Recognition

SAFE: Scale Aware Feature Encoder for Scene Text Recognition SAFE: Scale Aware Feature Encoder for Scene Text Recognition Wei Liu, Chaofeng Chen, and Kwan-Yee K. Wong Department of Computer Science, The University of Hong Kong {wliu, cfchen, kykwong}@cs.hku.hk arxiv:1901.05770v1

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

Color Local Texture Features Based Face Recognition

Color Local Texture Features Based Face Recognition Color Local Texture Features Based Face Recognition Priyanka V. Bankar Department of Electronics and Communication Engineering SKN Sinhgad College of Engineering, Korti, Pandharpur, Maharashtra, India

More information

WITH increasing penetration of portable multimedia. A Convolutional Neural Network Based Chinese Text Detection Algorithm via Text Structure Modeling

WITH increasing penetration of portable multimedia. A Convolutional Neural Network Based Chinese Text Detection Algorithm via Text Structure Modeling 1 A Convolutional Neural Network Based Chinese Text Detection Algorithm via Text Structure Modeling Xiaohang Ren, Yi Zhou, Jianhua He, Senior Member, IEEE, Kai Chen Member, IEEE, Xiaokang Yang, Senior

More information

Selection of Scale-Invariant Parts for Object Class Recognition

Selection of Scale-Invariant Parts for Object Class Recognition Selection of Scale-Invariant Parts for Object Class Recognition Gy. Dorkó and C. Schmid INRIA Rhône-Alpes, GRAVIR-CNRS 655, av. de l Europe, 3833 Montbonnot, France fdorko,schmidg@inrialpes.fr Abstract

More information

A Text Detection System for Natural Scenes with Convolutional Feature Learning and Cascaded Classification

A Text Detection System for Natural Scenes with Convolutional Feature Learning and Cascaded Classification A Text Detection System for Natural Scenes with Convolutional Feature Learning and Cascaded Classification Siyu Zhu Center for Imaging Science Rochester Institute of Technology, NY, USA zhu siyu@hotmail.com

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