A novel image compression technology based on vector quantisation and linear regression prediction. Issue 2 (3rd April 2021)

• Record Type: Journal Article
• Authors: Xu, Shuying; Chang, Chin-Chen; Liu, Yanjun
• Is Part Of: Connection science. Volume 33:Issue 2(2021)
• Page Start: 219
• Page End: 236
• Publication Date: 2021-04-03
• Subjects: Vector quantisation (VQ) -- pixel prediction -- linear regression -- compression ratio -- image quality; Neural computers -- Periodicals; Artificial intelligence -- Periodicals; Cognitive science -- Periodicals; Connectionism -- Periodicals; 006.3
• DOI: 10.1080/09540091.2020.1806206
• Languages: English
• ISSNs: 0954-0091

Abstract

Abstract : In the information age, a digital image is an important media for people's daily interactions. Looking to maintain the quality of the restored image, how to maximise the compression of images has become a challenging topic. Vector quantisation (VQ) compression is an easy-operating image compression method that can compress images to 1/16th of the original size. Based on VQ compresses, a novel image compression method is proposed in this paper. The proposed scheme compresses the image depending on the result of linear regression prediction which can significantly increase the compression ratio.