Viewing Distance-Based Perceived Error Control for Local Backlight Dimming

This paper presents an effective local dimming algorithm that preserves the perceived image quality consistently in the backlight dimmed images regardless of viewing distances. The proposed algorithm determines the clipped points of each LED block by limiting the maximum size of image area to be distorted, and guarantying the minimum image quality in terms of the peak signal-to-noise ratio. It also preserves the perceived image quality of the dimmed images by adjusting the size of area to be distorted depending on the viewing distances of users. Simulation results show that the proposed method improves the average PSNR and the average SSIM values of the dimmed images by 4.19-13.21 dB and 0.002-0.004, compared with the benchmark methods, for viewing distances of 2.5-6.5 m, respectively. The proposed method also maintains the image quality of the dimmed images regardless of the image's characteristics, compared with the benchmark methods. In addition, it is confirmed that the proposed method achieves the higher power savings as the viewing distance increases. At the viewing distance of 6.5 m, although the image quality of the dimmed images are higher than that of the benchmark methods, the proposed method consumes 0.56% less power.