The In-Depth Analysis of the Lossy Mode of the Next-Generation Image Compression Standard JPEG XL
Enes Alperen Buğaz, Emin Ölmez, Aysun Tasyapi Celebi, Orhan Akbulut · 2025
Developments in digital imaging technologies and the limitations of the traditional JPEG standard in meeting modern needs, such as high dynamic range and wide color gamut, have created a demand for a more efficient image compression format aimed at reducing file size while preserving visual quality. In this study, the architecture of the JPEG XL standard is examined in detail through its reference source codes, focusing on its core technologies like the perception- enhancing XYB color space, variable-size transforms, adaptive quantization, and in-loop filters. This analysis reveals the fundamental structure behind JPEG XL's high compression efficiency, which demonstrates superior performance on both PSNR and SSIM quality metrics compared to JPEG and WebP, particularly at lower bit rates. Furthermore, this study serves as a guide for research in the image compression field, especially within the national domain, by presenting a detailed architectural analysis.