Robust Model for Text Extraction from Complex Video Inputs Based on SUSAN Contour Detection and Fuzzy C Means Clustering
The proposed system introduces a novel approach for extracting text effectively from different types of complex video inputs. The valuable information within the text can be deployed for text indexing and localization. The proposed system uses contour based protocol like SUSAN algorithm for evaluating the contour detection. The system then explores candidate text area and refines the edges by Fuzzy C Means Clustering. The unwanted non-text portions are removed using morphological operation like dilation. The results obtained from the proposed implementation are then compared with the traditional algorithm used in prior research work for evaluating its efficiency. The result achieved outperforms all the prior algorithms for extracting text from different types of complex video input.
Keywords: Text Extraction, SUSAN, Fuzzy C Means Clustering, Morphological Operations
Download Full-Text
ABOUT THE AUTHOR
Narasimha Murthy K N
Mr. K N Narasimha Murthy is an experienced teacher for more than two decades and presently working as a Professor and HOD in department of Information Science and Engineering at City Engineering college Bangalore, Affiliated to Visvesvaraya Technological University, India. He got BE degree in 1989, M.Tech degree in 1993, MCA degree in 2003 and M.Tech degree in 2006. His areas of interest are Algorithms, Image processing and Computer Vision and also a member of MIE, MISTE, IACSIT.
Narasimha Murthy K N
Mr. K N Narasimha Murthy is an experienced teacher for more than two decades and presently working as a Professor and HOD in department of Information Science and Engineering at City Engineering college Bangalore, Affiliated to Visvesvaraya Technological University, India. He got BE degree in 1989, M.Tech degree in 1993, MCA degree in 2003 and M.Tech degree in 2006. His areas of interest are Algorithms, Image processing and Computer Vision and also a member of MIE, MISTE, IACSIT.