07072006

Title : Estimation of orientation of an inclined planar texture surface using DWT: An approach to Shape from Texture

Speaker : Prof. Sukhendu Das

Gledden Senior Visiting Fellow

Department of Computer Science and Engineering

Indian Institute of Technology, Madras

Time : Friday 11 am    July 7 2006     Room 1.24

Abstract :

Analysis of texture surfaces is a vital stage of processing necessary to extract information of object shapes from 3D Scenes – the area popularly known as “Shape from Texture” (SFT). The work presents a method based on wavelet-based dyadic decomposition and m-channel filtering to analyze texture characteristics of a 3D surface, for extracting the orientation of a planar surface. Analytical expression relating the surface geometry (depth and orientation) with 3D-texture profile, termed as the “Spatial Frequency Distribution" (SFD), has been formulated. The imaging geometry considered provides the following two significant advantages:

o Geometric separable analysis of the image signal along orthogonal directions to obtain the orientation

parameters and

o Error in estimation of orientation in one direction does not effect the estimation in the other direction.

A computationally efficient method based on DWT has been proposed to obtain texture profiles (called Scalograms). Scalograms in the sub-bands of wavelet decomposition are processed to generate Scalemaps, from which the orientation is computed using a closed form solution based on SFD. Results have been obtained for a set of simulated and real-world images. Error analysis and a method of 3D texture segmentation will also be presented.

Biography :

Dr. Sukhendu Das completed his B.Tech degree from IIT Kharagpur, from the Department of Electrical Engineering in 1985 and M. Tech Degree in the area of Computer Technology from IIT Delhi in 1987. He then obtained his Ph.D. degree from IIT Kharagpur in 1993. His current areas of research interests are: Visual Perception, Computational Intelligence, Computer Vision: Digital Image Processing and Pattern Recognition, Computer Graphics and Computational Science and Engineering. He has published extensively in these areas. He has received one best paper and a best design contest award. Dr. Das has been a faculty in the Deptt. Of CS&E, IIT Madras, INDIA, since 1989, where he is currently working as an Associate Professor. He has guided several PhD and M.S., M. Tech and B. Tech students. He had completed several international and national sponsored projects and consultancies, both as principle and co-investigators.