Non-cognitive Color and Texture Based Image Segmentation Amalgamation with Evidence Theory of crop images

The present scenario of image processing is approaching towards the perceptualization. This paper proposes perceptual segmentation with non cognitive low level color and texture features and the application is directly to crop images. Higher efficiency is guaranteed when human intervention is involved. This paper basically takes care of color texture based image segmentation specifically for the images in which the information frequencies are higher. Paper aims to present efficient and robust image segmentation of various crop images and providing some tuning between the low level color and texture feature with high level semantics to improve efficiency of segmentation. With the significant performance improvement a perceptual tuning can be used. Major difference between the normal image segmentation and perceptual image segmentation is also emphasis very clearly in this paper. The future work can be extended by involving non cognitive methodology such as evidence theory, data can be amalgam to make algorithm robust and efficient.