Performance Comparison of Typical Spatial Domain Image Enhancement for a Mobile Robot in Low Contrast Environment

Images could be understood and analyzed more easily through enhance processing. Generally, image enhancement can be divided into spatial domain enhancement and frequency domain enhancement according to different processing domains. Yet compared with frequency domain processing, spatial domain enhancement is more easily to realize real-time enhancement of images, namely, spatial domain processing is easier to satisfy the real-time requirements of applications in practical engineering. While, real-time is critical to robots for performing operations. Hence, this paper devotes to exploring performance of four typical spatial domain enhancement algorithms for a mobile robot operating in low contrast environment. More precisely, fundamental theory of contrast stretch, histogram equalization, histogram matching, and local histogram are overviewed. Furthermore, a mobile robot system is briefly introduced. Finally, in condition of low contrast, performance of above four algorithms are compared in mobile robot via experiments. This work will provide a basic reference for further research on image enhancement in robot vision.