Overlapped-Area Removal-Based Image Interpolation for Head-Mounted Displays

Abstract

In this paper, an overlapped-area removal-based image interpolation algorithm is proposed to significantly decrease the computation complexity when generating the barrel-distorted image for head-mounted display systems. Existing systems require high computation complexity for generating an output sequence with a high frame rate. To reduce the time complexity, the proposed algorithm uses the low-resolution image as an input, and increases the overall resolution based on an overlapped-area removal technique. Specifically, in the analysis process, it generates the coordinate mapping table after analyzing the overlapped area. In the synthesis process, the proposed algorithm performs the overlapped-area removal to reduce the computational complexity, and then it interpolates an input image. The results of the simulation show that the proposed algorithm maximally maintains the image quality while reducing the computation time greatly. Specifically, it reduces the average computation time up to 36.32% while improving peak signal-to-noise ratio by up to 3.72 dB, when compared to the benchmark algorithm.