Unveiling the Significance of Sign Calculation Impact on JPEG Applications

This paper aims to enhance the balance between compression efficiency and image quality in JPEG by employing an approximate adder with sign error correction techniques. Approximate computing is getting significant attention as a novel approach for enhancing software and hardware performance. Examples of applications that leverage approximate computing include data mining, recognition, audio processing, and web search. However, unfortunately, applying approximate calculations does not always guarantee an improvement in performance. This is because it is difficult for certain applications to achieve the anticipated performance due to the adverse impact of sign calculation errors. To address this problem, this paper applied an approximate adder that takes into account sign calculation errors. The compression ratio was successfully improved while maintaining the image quality. To the best of our knowledge, there are few studies on the application of approximate adders focusing on compression ratio.