Imran Siddiqi

Text independent writer recognition using redundant writing patterns with contour-based orientation and curvature features

We propose an effective method for automatic writer recognition from unconstrained handwritten text images. Our method relies on two different aspects of writing: the presence of redundant patterns in the writing and its visual attributes. Analyzing small writing fragments, we seek to extract the patterns that an individual employs frequently as he writes. We also exploit two important visual attributes of writing, orientation and curvature, by computing a set of features from writing samples at different levels of observation. Finally we combine the two facets of handwriting to characterize the writer of a handwritten sample. The proposed methodology evaluated on two different data sets exhibits promising results on writer identification and verification.

A Set of Chain Code Based Features for Writer Recognition

This communication presents an effective method for writer recognition in handwritten documents. We have introduced a set of features that are extracted from the contours of handwritten images at different observation levels. At the global level, we extract the histograms of the chain code, the first and second order differential chain codes and, the histogram of the curvature indices at each point of the contour of handwriting. At the local level, the handwritten text is divided into a large number of small adaptive windows and within each window the contribution of each of the eight directions (and their differentials) is counted in the corresponding histograms. Two writings are then compared by computing the distances between their respective histograms. The system trained and tested on two different data sets of 650 and 225 writers respectively, exhibited promising results on writer identification and verification.

Writer Identification in Handwritten Documents

This work presents an effective method for writer identification in handwritten documents. We have developed a local approach, based on the extraction of characteristics that are specific to a writer. To exploit the existence of redundant patterns within a handwriting, the writing is divided into a large number of small sub-images, and the sub-images that are morphologically similar are grouped together in the same classes. The patterns, which occur frequently for a writer are thus extracted. The author of the unknown document is then identified by a Bayesian classifier. The system trained and tested on 50 documents of the same number of authors, reported an identification rate of 94%.

Text-independent writer recognition using multi-script handwritten texts

This paper presents a text-independent writer recognition method in a multi-script environment. Handwritten texts in Greek and English are considered in this study. The objective is to recognize the writer of a handwritten text in one script from the samples of the same writer in another script and hence validate the hypothesis that writing style of an individual remains constant across different scripts. Another interesting aspect of our study is the use of short handwritten texts which was implied to resemble the real life scenarios where the forensic experts, in general, find only short pieces of texts to identify a given writer. The proposed method is based on a set of run-length features which are compared with the well-known state-of-the-art features. Classification is carried out using K-Nearest Neighbors (K-NN) and Support Vector Machines (SVM). The experimental results obtained on a database of 126 writers with 4 samples per writer show that the proposed scheme achieves interesting performances on writer identification and verification in a multi-script environment.

Writer identification using texture descriptors of handwritten fragments

Abstract This paper presents a texture based approach for identification of writers from offline images of handwriting. Contrary to the classical texture based techniques which extract texture information at page or block level, we exploit the texture at a very small observation scale. The proposed technique divides a given handwriting into small fragments and considers each fragment as a texture. Texture descriptors including histograms of Local Binary Patterns (LBP), Local Ternary Patterns (LTP) and Local Phase Quantization (LPQ) are then computed from these fragments. The writer of a document is characterized by the set of histograms calculated from all the fragments in the writing. Two writings are compared by computing the distance between the descriptors of their writing fragments. The technique evaluated on IFN/ENIT and IAM databases comprising handwritten text in Arabic and English, respectively, realized high identification rates.

Automatic analysis of handwriting for gender classification

This paper presents a study to predict gender of individuals from scanned images of their handwritings. The proposed methodology is based on extracting a set of features from writing samples of male and female writers and training classifiers to learn to discriminate between the two. Writing attributes like slant, curvature, texture and legibility are estimated by computing local and global features. Classification is carried out using artificial neural networks and support vector machine. The proposed technique evaluated on two databases under a number of scenarios realized interesting results on predicting gender from handwriting.

Towards Searchable Digital Urdu Libraries - A Word Spotting Based Retrieval Approach

Libraries in South Asia hold huge collections of valuable printed documents in Urdu and it is of interest to digitize these collections to make them more accessible. The unavailability of an OCR for Urdu however limits the concept of a digital Urdu library to scanning of documents only, offering very limited search facility based on manually assigned tags. We address this issue by proposing a word spotting based keyword search method for information retrieval in digitized collections of printed Urdu documents. The proposed method is based on segmentation of Urdu text in to partial words and representing each partial word by a set of features. To search a specific word (or phrase), the user provides a query in the form of an image. Comparing the features of the partial words in the query image with the ones already indexed, the user is provided with a list of documents containing occurrences of the queried word. The system evaluated on 50 Urdu documents exhibited a recall of 95.17% and a precision of 94.3%.

Edge-Based Features for Localization of Artificial Urdu Text in Video Images

Content-based video indexing and retrieval has become an interesting research area with the tremendous growth in the amount of digital media. In addition to the audio-visual content, text appearing in videos can serve as a powerful tool for semantic indexing and retrieval of videos. This paper proposes a method based on edge-features for horizontally aligned artificial Urdu text detection from video images. The system exploits edge based segmentation to extract textual content from videos. We first find the vertical gradients in the input video image and average the gradient magnitude in a fixed neighborhood of each pixel. The resulting image is binarized and the horizontal run length smoothing algorithm (RLSA) is applied to merge possible text regions. An edge density filter is then applied to eliminate noisy non-text regions. Finally, the candidate regions satisfying certain geometrical constraints are accepted as text regions. The proposed approach evaluated on a data set of 150 video images exhibited promising results.

Offline cursive Urdu-Nastaliq script recognition using multidimensional recurrent neural networks

Optical Character Recognition of cursive scripts remains a challenging task due to a large number of character shapes, inter- and intra-word overlaps, context sensitivity and diagonality of text. This paper presents an implicit segmentation based recognition system for Urdu text lines in Nastaliq script. The proposed technique relies on sliding overlapped windows on lines of text and extracting a set of statistical features. The extracted features are fed to a multi-dimensional long short term memory recurrent neural network (MDLSTM RNN) with a connectionist temporal classification (CTC) output layer that labels the character sequences. Experimental study of the proposed technique is carried out on the standard Urdu Printed Text-line Images (UPTI) database which comprises 10,000 text lines in Nastaliq font. Evaluations under different experimental settings realize promising recognition rates with a highest character recognition rate of 96.40%.

Detection and Recognition of Traffic Signs from Road Scene Images

Automatic detection and recognition of road signs is an important component of automated driver assistance systems contributing to the safety of the drivers, pedestrians and vehicles. Despite significant research, the problem of detecting and recognizing road signs still remains challenging due to varying lighting conditions, complex backgrounds and different viewing angles. We present an effective and efficient method for detection and recognition of traffic signs from images. Detection is carried out by performing color based segmentation followed by application of Hough transform to find circles, triangles or rectangles. Recognition is carried out using three state-of-the-art feature matching techniques, SIFT, SURF and BRISK. The proposed system evaluated on a custom developed dataset reported promising detection and recognition results. A comparative analysis of the three descriptors reveal that while SIFT achieves the best recognition rates, BRISK is the most efficient of the three descriptors in terms of computation time.