Noise-aware camera exposure control for robust robot vision

Abstract

In this thesis, we propose a noise-aware exposure control algorithm for robust robot vision. Our method aims to capture the best-exposed image which can boost the performance of various computer vision and robotics tasks. For this purpose, we carefully design an image quality metric which captures complementary quality attributes and ensures light-weight computation. Specifically, our metric consists of a combination of image gradient, entropy, and noise metrics to capture the appreciate property of an image. The synergy of these measures allows preserving sharp edge and rich texture in the image while maintaining a low noise level. Using this novel metric, we propose a real-time and fully automatic exposure and gain control technique based on the Nelder-Mead(NM) Simplex method. Also, we proposed a Modifed NM method that can react fastly and robustly to the luminance change of dynamic environment. To illustrate the effectiveness of our technique, a large set of experimental results, including feature matching, pose estimation, object detection, disparity estimation, and real-world experiments, demonstrates higher qualitative and quantitative performances when compared with conventional approaches.