Implementation of enhanced fractal texture analysis with layout descriptor for image segmentation

Extraction of image features can be in various forms based on color, shape and the texture. The identification and segregation of the objects, a difficult task and more challenging because of many reasons like the variations in the background which is not stationary, illumination changes etc., This article aims at studying and analyzing feature extraction and object recognition. The hybrid approach that is formed by combining enhanced Fractal Texture Analysis with Layout descriptor is capable of meeting the issues in image segmentation. The help of designed shape and texture based Image segmentation does implementation. The image is pre-processed in this process in the beginning; a noise protected edge detector is used. At last, morphological gradient with both shape and texture gradient are applied. The experimental results depict the robustness of this method. The help of CV online object database that contains varied set of object images validates this approach.