Modified Firefly Model-Based Vector Quantization For Clinical Medical Image Compression

Abstract

Due to the rapid increase in the usage of medical images for disease diagnosis and the rise in the volume of data produced by different medical imaging equipment, the transmission and archival of images need data compression. In the past decade, various image compression methods have been presented and find its applicability in various fields. Vector Quantization (VQ) plays a vital part in compressing images, and a Quantization Table (QT) construction is a significant process. The effectiveness of any compression technique mainly relies on the QT, generally a matrix of 64 integers. Selecting a QT is an optimization issue that bio-inspired techniques can address. The article compares two QT selection algorithms: Firefly with the Tumbling effect (FF-Tumbling) and the Teaching and Learning Based Optimization (FF-TLBO) approach. An extensive study is made between these two methods and analyzed the results. The simulation outcome is interesting in that the FF-Tumbling approach can achieve optimal reconstructed image quality, and the FF-TLBO method has the efficiency to achieve optimal compression performance. © 2023 Seventh Sense Research Group