A Hybrid Approach for Region-Based Medical Image Compression with Nature-Inspired Optimization Algorithm

Autor: D. Sujitha Juliet, S. Saravanan
Rok vydání: 2021
Předmět:
Zdroj: Innovations in Computer Science and Engineering ISBN: 9789813345423
DOI: 10.1007/978-981-33-4543-0_24
Popis: Medical modalities generate a massive amount of digital volumetric data to analyze and diagnose medical problems. To sure over the data quality and storage space, compression primes to be an efficient methodology in the digital world. Medical images represent the body features for a diagnostic purpose that needs to get compressed without a loss of information. Reducing the redundancies and representation in a shorter manner achieved over a region of interest area on an image solves the problem. The proposed methodology uses the region-based active contour method driven by bat algorithm to segment an image into a region of interest and non-region of interest. Region of interest area compressed by a lossless integer-based Karhunen-Loeve transforms, where the non-region of interest compressed by a lossy Karhunen-Loeve transforms. Optimum results suggest that the proposed method improvises the region segmentation of a medical image, which results in achieving a high PSNR, SSIM and a quality compressed image.
Databáze: OpenAIRE