An Efficient Hybrid Optimization Algorithm for Image Compression

Abstract

In this work, a novel image compression approach is developed that is processed in several series of technologies. Here, the first process is the image segmentation and it is done using Adaptive ACM that partitions or segments the image into two regions such as ROI as well as nonROI. Here, the adaptiveness of this ACM is determined with the idea of optimization algorithm. To handle the ROI regions, the JPEG-LS technique is exploited and to handle the non-ROI region the wavelet-based lossy compression technique is utilized. The outcome of both the JPEG-LS technique, as well as a wavelet-based compression approach is integrated with respect to the bit-stream amalgamation in order to produce the compressed image. Then, the compressed image is exploited to the image decompression that will be the overturn process of compression. It will comprise the bitstream separation that is subsequently individually process in both the wavelet-based decomposition and JPEG-LS decoding for obtaining the non-ROI regions and ROI. At last, the original image is obtained accurately. Moreover, the main objective of this paper falls in the adaptiveness under optimization. The maximum iteration and weighting factor in ACM are optimally chosen and for this a novel hybrid optimization technique is proposed, which hybridizes the concept of Differential Evolution method with Monarch Butterfly Optimization Algorithm. Here, the proposed method is compared with the conventional methods in order to shows its effectiveness for image compression.