Performance analysis of modified wavelet difference reduction methods in image compression and transmission

Abstract

The wavelet difference reduction (WDR) method, a variant of run-length coding, finds its significance in data transmission applications. Over time, numerous enhanced iterations of WDR methods have emerged. Notably, the Adaptive Scalable WDR method exhibits superior coding gains, as evidenced by the peak signal-to-noise ratio (PSNR) and structural similarity index metric (SSIM), when compared to its predecessors. This paper conducts an exhaustive examination, encompassing both coding performance and time complexity, of various WDR methods vis-à-vis the conventional image compression algorithm SPIHT. Furthermore, it delves into the performance assessment of diverse coding techniques in the realm of encoding arbitrary-shaped objects. The analysis underscores that modified WDR variants demonstrate remarkable prowess in compression, rendering them invaluable for rapid transmission in bandwidth-constrained networks. To substantiate these findings, coding results (measured in terms of PSNR) are derived from the application of these algorithms to standard test images, MRI images, and video still images. The results reveal coding gains ranging from 0.5 dB to 1 dB for regular resolution images and a substantial 2 dB to 12 dB for scalable resolution scenarios, in comparison to traditional coding approaches. Consequently, this analysis underscores the convenience and superiority of modified WDR methods, not only for still images but also for encoding and transmitting arbitrary-shaped objects.

Comparative Analysis of Wavelet Families in Image Compression, Featuring the Proposed New Wavelet

“…The evaluation involves assessing performance using Peak Signal-to-Noise Ratio (PSNR) and Mean Square Error (MSE). The results highlight differences among wavelet families [22].…”

“…The zerosum area beneath the wavelet curve ensures equal energy distribution in both positive and negative directions. In signal processing, wavelets effectively recover weak signals from noise and analyze signals with dynamic frequency changes over time [22].…”