Title :
Fast cellular feature extraction method
Author :
Lai, Kwok-Kei ; Wu, Paul S Y
Author_Institution :
Dept. of Manuf. Eng., City Univ. of Hong Kong, Kowloon, Hong Kong
Abstract :
A novel and speedy feature extraction method, called cellular feature extraction method (CFEM), is proposed in this paper. The feature extraction process is particularly important in image processing because it aims to identify all target pixels of the image. The proposed CFEM, inter alia, is particularly effective in extracting useful features of line drawings and more time-saving in processing larger-size images. The main idea behind CFEM is to ignore most of the non-target pixels which bear no useful information to the job at-hand. By design, the CFEM could selectively read out all target pixels of an image. As such, the CFEM can also act as a preprocessing method, and is fully compatible with other image processing techniques, such as filtering and thinning. When CFEM is combined with other image processing techniques to process an image, same results will be obtained in lesser time than that without CFEM. The results of experimenting CFEM in line drawings show a time-reduction of 74% in processing the largest testing image. The multiresolution properties and the limitations of CFEM are discussed
Keywords :
feature extraction; image resolution; cellular feature extraction method; filtering; image processing; line drawings; multiresolution properties; preprocessing method; thinning; Computer aided manufacturing; Digital images; Engineering drawings; Feature extraction; Hardware; Image processing; Image resolution; Object recognition; Pixel; Pulp manufacturing;
Conference_Titel :
Systems, Man, and Cybernetics, 1996., IEEE International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-3280-6
DOI :
10.1109/ICSMC.1996.569777