A super-pixel based on-chip compression is proposed in this paper. The compression is achieved by reading only one sample for each super-pixel. The proposed technique and the corresponding circuit are simulated in MATLAB and UMC 180 nm CMOS technology, respectively. Higher values of PSNR are observed as compared to the state-of-the-art on-chip compression techniques. For the compression factor of 2, the implemented design results in 49% of power saving.
Zhongxiang CaoYangfan ZhouQuanliang LiQi QinLiyuan LiuNanjian Wu
Yukinari NishikawaShoji KawahitoMasanori FurutaToshihiro Tamura
Yukinari NishikawaShoji KawahitoMasanori FurutaToshihiro Tamura
Yukinari NishikawaShoji KawahitoMasanori FurutaToshihiro Tamura