Welcome to the InfoSci Platform
Feature Extraction Algorithms to Color Image

Feature Extraction Algorithms to Color Image

QingE Wu, Weidong Yang
Copyright: © 2018 |Pages: 24
ISBN13: 9781522552048|ISBN10: 1522552049|EISBN13: 9781522552055
DOI: 10.4018/978-1-5225-5204-8.ch016
Cite Chapter Cite Chapter

MLA

Wu, QingE, and Weidong Yang. "Feature Extraction Algorithms to Color Image." Computer Vision: Concepts, Methodologies, Tools, and Applications, edited by Information Resources Management Association, IGI Global, 2018, pp. 397-420. https://doi.org/10.4018/978-1-5225-5204-8.ch016

APA

Wu, Q. & Yang, W. (2018). Feature Extraction Algorithms to Color Image. In I. Management Association (Ed.), Computer Vision: Concepts, Methodologies, Tools, and Applications (pp. 397-420). IGI Global. https://doi.org/10.4018/978-1-5225-5204-8.ch016

Chicago

Wu, QingE, and Weidong Yang. "Feature Extraction Algorithms to Color Image." In Computer Vision: Concepts, Methodologies, Tools, and Applications, edited by Information Resources Management Association, 397-420. Hershey, PA: IGI Global, 2018. https://doi.org/10.4018/978-1-5225-5204-8.ch016

Export Reference

Mendeley
Favorite

Abstract

The existing image processing algorithms mainly studied on feature extraction of gray image with one-dimensional parameter, such as edges, corners. However, the extraction of some characteristic points to color image with three-dimensional parameters, such as the extraction of color edge, corner points, inflection points, etc., is an image problem to be urgently solved. In order to carry out a fast and accurate feature extraction on color image, this paper proposes two types of extraction algorithms to color edge and corner points of color image, i.e., similar color segment algorithm and pixel probabilistic algorithm, compares with the two algorithms, gives the two algorithms are used to different color distribution situations, as well as shows the extraction effect of color by the combination of the two algorithms, moreover, gives the contrast experiment and effect analysis of the two algorithms. To compare the similar color segment algorithm with the probabilistic algorithm, experimental results show that the similar color segment algorithm is better than the pixel probabilistic algorithm under the more obvious color edge, because it has the better edge detection, stronger anti-noise ability, faster processing speed and other advantages. Under the transition phase of color edge is gentle or color edge is no clear, the image detection effect of the pixel probabilistic algorithm is better than that of the similar color segment algorithm. But the combinative effect of the two algorithms is the best in this case, which is more close to the color effect of original image. Moreover, this paper analyzes the performance of the similar color segment algorithm, and gives the comparison of the proposed two algorithms and existing classical algorithms used usually to feature extraction of color image. The two algorithms proposed and these researches development in this paper have not only enriched the contents of image processing algorithms, but also provide a solution tool for image segmentation, feature extraction to target, precise positioning, etc., such as extraction of complexion, physiological color photographs processing, feature extraction of ionosphere, detection and extraction of biological composition of oceans, to be applied to a lots of departments, such as the police, hospital departments, surgery, polar department, and so on, as well as provide a way of thinking for the rapid, accurate detection of case, surgery, scientific research information search.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.