Unifying Dual-Pyramid Structure and Y–G Channel Synergy for Full-Reference Image Quality Assessment

Ali Abdulazeez Mohammed Baqer Qazzaz
Yousif Samer Mudhafar
Siraj Muneer Mahboba

Advance Sustainable Science, Engineering and Technology (ASSET)
Vol. 8 No. 1 (2026): November - January

Standard metrics such as SSIM often overlook complex chromatic distortions, creating a gap between objective scores and human judgment. To address this, we present the Synergistic Structural Similarity Index (SSSI), a metric grounded in a novel dual-pyramid strategy that integrates Gaussian-blurred stability with direct subsampling sharpness. Our method departs from luminance-only analysis by employing an equal, synergistic partnership between the luminance (Y) and Green (G) channels, mirroring the eye's spectral sensitivity. On the KADID-10k dataset, SSSI achieves an SROCC of 0.7793. This represents a significant 4% performance gain over the standard SSIM baseline, demonstrating that integrating chromatic data with dual-scale structural analysis provides a more accurate proxy for human visual perception.