Logo Retrieval using Morphological Pattern Spectrum

Abstract

In this paper we propose a geometric approach to capture the shape and spatial relation attributes in binary images to build an efficient logo retrieval system. We extract the shape feature by computing the morphological pattern spectrum (MPS) for various structuring elements. The spatial relation among the disjoint components in the logo image is described by defining a simple and computationally inexpensive spatial relation graph (SRG). The SRG is obtained by joining the centroids of the two largest components with the centroids of the rest of the components. For retrieval purposes, the similarity measure is defined with respect to similarities in both shape and spatial relationship (MPS and SRG). The joint shape and spatial relation features enable us to obtain a very good precision rate. A relevance feedback mechanism is used to obtain the weights for various components in the similarity measure. The details of the method as well as results of experimentations are presented.

Indexing terms:

• Shape
• Spatial
• Content Based Image Retrieval System (CBIR)
• Morphological Pattern Spectrum (MPS)
• Spatial Relation Graph (SRG)

Additional information

Notes on contributors

Rajashekhar

Rajashekhara did his BE in Electronics and Communication Engineering from Kuvempu University in 1994 and MTech in Electronics and Communication Systems from Mysore University in 1997. He is working as a Lecturer at National Institute of Technology, Surathkal, Karnataka. Currently, he is a PhD student in the Department of Electrical Engineering, IIT Bombay. His research interests include Digital Signal Processing, Image Processing and Computer Vision.

Subhasis Chaudhuri

Subhasis Chaudhuri received his MTech degree in Electronics and Electrical Communication Engineering from the Indian Institute of Technology, Kharagpur in 1985. He received the MS and the PhD degrees, both in Electrical Engineering, from the University of Calgary, Canada and the University of California, San Diego respectively. Currently, he is serving as a Professor in IIT Bombay. He is a Fellow of the Alexander von Humboldt Foundation, Germany, the Indian National Academy of Engineering, the National Academy of Sciences, and the Indian Academy of Sciences. He is the recipient of Dr. Vikram Sarabhai Research Award for the year 2001, and the Swarnajayanti Fellowship in 2003. He received SS Bhatnagar Prize in Engineering Sciences for the year 2004. He is the coauthor of the following books ‘Depth from Defocus: A Real Aperture Imaging Approach’, and ‘Motion-free Super Resolution’, both published by Springer, NY. He has also edited a book on ‘Super-resolution Imaging’ published by Kluwer Academic in 2001. His research interests include image processing, computer vision and multimedia.