An Approach of Hybrid Cuckoo Search with Hill Climbing (CS-HC) Algorithm for Segmentation based on Multilevel Thresholding and Enhanced Lossless Prediction based Compression Algorithm for DICOM CT Images

Vishal Goyal and Aasheesh Shukla

In applications of computer vision, a common image processing step is image segmentation. It is used segment the pixels into various classes. The increase in the threshold count will complicate the process of image segmentation. At the same time, it becomes a NT problem in the field of threshold application in image. This work proposes a, optimization technique based multilevel thresholding to extract ROI and uses improved lossless prediction algorithm to compress DICOM images in telemedicine applications. The mechanism used by search agent to update the optimum solution is enhanced by hybrid Cuckoo search with hill climbing (CS-HC) algorithm. Threshold value is estimated by this algorithm. The superior results are produced by the proposed CS-HC based multilevel level thresholding as shown by the results of simulation. The optimization is made effective and it has high convergence rate. When compared with JPEG lossless and lossy compression techniques, efficient results are produced by the proposed lossless compression algorithm based on classification and blending prediction. With different threshold values, the efficiency of the algorithm is tested. Matlab2010a is used to implement this algorithm and DICOM images are used to test it.

Volume 11 | 10-Special Issue

Pages: 633-641

DOI: 10.5373/JARDCS/V11SP10/20192852