Super-Resolution Reconstruction Algorithm for PMD Range Image Based on Regularization

Abstract

PMD camera is a three-dimensional imaging system based on TOF principle. In the basis on obtaining the two-dimensional gray image, this camera can capture range image and amplitude image at the same time. Their main drawbacks are low-resolution and random noise. According to this problem, this paper proposed an improved regularization method which uses adaptive weighting on the regular terms, and uses confidence as the weighted value of each pixel variance of range image to smooth each pixel in neighborhood. The experimental results demonstrate that this algorithm not only can effectively enhance the edge of range image, but also can reduce the noise and the performance of reconstruction is better improved.