Unhealthy Detection in Livestock Texture Images using Subsampled Contourlet Transform and SVM

In this paper a new split and merge algorithm based on Contourlet transform and Support Vector Machine (SVM) is presented for automatic segmentation and classification of unhealthy in Livestock Texture Images. We focused on the liver textural images of livestock to verify if there is any unhealthy on its textural image. The Contourlet transform is used because it allows analysis of images with various resolution levels and directions. It effectively captures smooth contours that are dominant features in textural images. In addition, we have used SVM classifier to classify the texture features. The proposed method provides a fast algorithm with enough accuracy that can be implemented in a parallel structure for realtime processing. The simulation results show the effectiveness of the new proposed algorithm.