arXiv subjects

Dipak Kumar Basu

Publications and source records attributed to Dipak Kumar Basu.

40 records · Page 3Linked to original sources

arXiv2010

Classification of Polar-Thermal Eigenfaces using Multilayer Perceptron for Human Face Recognition

This paper presents a novel approach to handle the challenges of face recognition. In this work thermal face images are considered, which minimizes the affect of illumination changes and occlusion due to moustache, beards, adornments etc. The proposed approach registers the training and testing thermal face images in polar coordinate, which is capable to handle complicacies introduced by scaling and rotation. Polar images are projected into eigenspace and finally classified using a multi-layer perceptron. In the experiments we have used Object Tracking and Classification Beyond Visible Spectrum (OTCBVS) database benchmark thermal face images. Experimental results show that the proposed approach significantly improves the verification and identification performance and the success rate is 97.05%.

cs.CV↗

arXiv2010

Reduction of Feature Vectors Using Rough Set Theory for Human Face Recognition

In this paper we describe a procedure to reduce the size of the input feature vector. A complex pattern recognition problem like face recognition involves huge dimension of input feature vector. To reduce that dimension here we have used eigenspace projection (also called as Principal Component Analysis), which is basically transformation of space. To reduce further we have applied feature selection method to select indispensable features, which will remain in the final feature vectors. Features those are not selected are removed from the final feature vector considering them as redundant or superfluous. For selection of features we have used the concept of reduct and core from rough set theory. This method has shown very good performance. It is worth to mention that in some cases the recognition rate increases with the decrease in the feature vector dimension.

cs.CV↗

arXiv2010

Text/Graphics Separation for Business Card Images for Mobile Devices

Separation of the text regions from background texture and graphics is an important step of any optical character recognition sytem for the images containg both texts and graphics. In this paper, we have presented a novel text/graphics separation technique for business card images captured with a cell-phone camera. At first, the background is eliminated at a coarse level based on intensity variance. This makes the foreground components distinct from each other. Then the non-text components are removed using various characteristic features of text and graphics. Finally, the text regions are skew corrected and binarized for further processing. Experimenting with business card images of various resolutions, we have found an optimum performance of 98.54% with 0.75 MP images, that takes 0.17 seconds processing time and 1.1 MB peak memory on a moderately powerful computer (DualCore 1.73 GHz Processor, 1 GB RAM, 1 MB L2 Cache). The developed technique is computationally efficient and consumes low memory so as to be applicable on mobile devices.

cs.GR↗

arXiv2010

Word level Script Identification from Bangla and Devanagri Handwritten Texts mixed with Roman Script

India is a multi-lingual country where Roman script is often used alongside different Indic scripts in a text document. To develop a script specific handwritten Optical Character Recognition (OCR) system, it is therefore necessary to identify the scripts of handwritten text correctly. In this paper, we present a system, which automatically separates the scripts of handwritten words from a document, written in Bangla or Devanagri mixed with Roman scripts. In this script separation technique, we first, extract the text lines and words from document pages using a script independent Neighboring Component Analysis technique. Then we have designed a Multi Layer Perceptron (MLP) based classifier for script separation, trained with 8 different wordlevel holistic features. Two equal sized datasets, one with Bangla and Roman scripts and the other with Devanagri and Roman scripts, are prepared for the system evaluation. On respective independent text samples, word-level script identification accuracies of 99.29% and 98.43% are achieved.

cs.LG↗