期刊名称：International Journal of Signal Processing, Image Processing and Pattern Recognition
印刷版ISSN：2005-4254
出版年度：2016
卷号：9
期号：6
页码：71-88
DOI：10.14257/ijsip.2016.9.6.07
出版社：SERSC
摘要：Color constancy allows visitors to recognize color in the diversity of conditions, and to see some regularity in the color. Generally, color constancy is a method of which calculates the particular impact involving discrete light sources using a digital image. The particular image documented with a camera is resolute by several components: the actual physical information with the picture, the particular illumination occurrence within the scene, as well as the features with the digital camera. Color based images have several applications. Retrieval based on object colors must take into an explanation the features that manipulate the pattern of color images: illumination conditions, sensor spectral sensibilities and surface reflectance. Tracking objects in environments under non-uniform lighting condition is mainly challenging as the practical appearance may alter in space and time [1]. We create an integrated effort of fuzzy membership with edge preserving filtering. As fuzzy membership exploits two classes because we make use of dual membership functions in the research work. In this paper, improved results have extracted for color constancy from optimal method based on fuzzy membership with edge preservation filtering. An evaluation of the proposed technique is also drawn with existing techniques, the comparisons have obviously shown that the fuzzy based color constancy outperforms over the available techniques.
关键词：Fuzzy membership; edge preserving filtering; gray world; color constancy
