Work place: Department of Computer Science, Jadavpur University, Kolkata, India
Research Interests: Image Processing, Machine Learning, Artificial Intelligence, Biometrics
Debotosh Bhattacharjee received the MCSE and Ph. D.(Eng.) degrees from Jadavpur University, India, in 1997 and 2004 respectively. He was associated with different institutes in various capacities until March 2007. After that he joined his Alma Mater, Jadavpur University. His research interests pertain to the applications of computational intelligence techniques like Fuzzy logic, Artificial Neural Network, Genetic Algorithm, Rough Set Theory, Cellular Automata etc. in Face Recognition, OCR, and Information Security. He is a life member of Indian Society for Technical Education (ISTE, New Delhi), Indian Unit for Pattern Recognition and Artificial Intelligence (IUPRAI), and senior member of IEEE (USA).
DOI: https://doi.org/10.5815/ijigsp.2015.05.03, Pub. Date: 8 Apr. 2015
In this paper, authors have proposed two novel techniques for occlusion detection and then localization of the occluded section from a given 3D face image if occlusion is present. For both of these methods, at first, the 2.5D or range face images are created from input 3D face images. Then for detecting the occluded faces, two approaches have been followed, namely: block based and threshold based. These two methods have been investigated individually on Bosphorus database for localization of occluded portion. Bosphorus database consists of different types of occlusions, which have been considered during our research work. If 2D and 3D images are compared then 3D images provide more reliable, accurate, valid information within digitized data. In case of 2D images each point, named as pixel, is represented by a single value. One byte for gray scale and three byte for color images in a 2D grid whereas in case of 3D, there is no concept of 2D grid. Each point is represented by three values, namely X, Y and Z. The 'Z' value in X-Y plane does not contain the reflected light energy like 2D images. The facial surface's depth data is included in Z's point set. The threshold or cutoff based technique can detect the occluded faces with the accuracy 91.79% and second approach i.e. block based approach can successfully detect the same with the success rate of 99.71%. The accuracy of the proposed occlusion detection scheme has been measured as a qualitative parameter based on subjective fidelity criteria.[...] Read more.
DOI: https://doi.org/10.5815/ijigsp.2015.02.05, Pub. Date: 8 Jan. 2015
This paper presents a gait recognition method which is based on spatio-temporal movement characteristics of human subject with respect to surveillance camera. Different measures, like leg rise from ground (LRFG), the angles created between the legs with the centre of Mass (ABLC), angles created between the Centre of Mass Knee and Ankle with the (CKA), angles created between Centre of Mass, Wrist and knee (CWK), the distances between the control points and centre of Mass (DCC) have been taken as different features. Fourier descriptor has been used for shape extraction of individual frames of a subject. Statistical approach has been used for recognition of individuals based on the n feature vectors, each of size 23(collected from LRFG, ABLCs, CKA, CWK and DCCs) for each video frame. It has been found that recognition result of our approach is encouraging with compared to other recent methods.[...] Read more.
DOI: https://doi.org/10.5815/ijigsp.2014.07.03, Pub. Date: 8 Jun. 2014
In this work, a simple characterization of human gait, which can be used for surveillance purpose, is presented. Different measures, like leg rise from ground (LRFG), the angles created between the legs with the centroid (ABLC), the distances between the control points and centroid (DBCC) have been taken as different features. In this method, the corner points from the edge of the object in the image have been considered. Out of several corner points thus extracted, a set of eleven significant points, termed as control points, that effectively and rightly characterize the gait pattern, have been selected. The boundary of the object has been considered and using control points on the boundary the centroid of those has been found out. Statistical approach has been used for recognition of individuals based on the n feature vectors, each of size 23(collected from LRFG, ABLCs, and DBCCs) for each video frame, where n is the number of video frames in each gait cycles. It has been found that recognition result of our approach is encouraging with compared to other recent methods.[...] Read more.
DOI: https://doi.org/10.5815/ijitcs.2013.09.03, Pub. Date: 8 Aug. 2013
This paper exploits the feature extraction capabilities of the discrete cosine transform (DCT) together with an illumination normalization approach in the logarithm domain that increase its robustness to variations in facial geometry and illumination. Secondly in the same domain the entropy measures are applied on the DCT coefficients so that maximum entropy preserving pixels can be extracted as the feature vector. Thus the informative features of a face can be extracted in a low dimensional space. Finally, the kernel entropy component analysis (KECA) with an extension of arc cosine kernels is applied on the extracted DCT coefficients that contribute most to the entropy estimate to obtain only those real kernel ECA eigenvectors that are associated with eigenvalues having high positive entropy contribution. The resulting system was successfully tested on real image sequences and is robust to significant partial occlusion and illumination changes, validated with the experiments on the FERET, AR, FRAV2D and ORL face databases. Experimental comparison is demonstrated to prove the superiority of the proposed approach in respect to recognition accuracy. Using specificity and sensitivity we find that the best is achieved when Renyi entropy is applied on the DCT coefficients. Extensive experimental comparison is demonstrated to prove the superiority of the proposed approach in respect to recognition accuracy. Moreover, the proposed approach is very simple, computationally fast and can be implemented in any real-time face recognition system.[...] Read more.
DOI: https://doi.org/10.5815/ijigsp.2012.02.05, Pub. Date: 8 Mar. 2012
Recently human gait has become a promising and very important biometric for identification. Current research on gait recognition is usually based on an average gait image or a silhouette sequence, or a motion structure model. In this paper, the information about gait is obtained from the disparity on time and space of the different parts of the silhouette. This paper proposes a gait recognition method using edge detection, identification of corner points from edges, and selection of control points out of those corner points. Here, the images of moving human figures are subtracted from background by simple background modeling technique to obtain binary silhouettes. A gait signature of a person is taken as silhouette images of a complete gait cycle. A complete gait cycle is then divided into different frames in such a way that the information of the person’s gait style can be represented fully. One given unknown gait cycle is compared with stored gait cycles in terms of a cyclic distances between control points of an image of input gait cycle with that of corresponding image of the stored gait cycle. Experimental results show that our method is encouraging in terms of recognition accuracy.[...] Read more.
Subscribe to receive issue release notifications and newsletters from MECS Press journals