Vision System Actualization for Object Localization Robot

The accuracy of object localization by an autonomous robotic system depends on the efficiency of the vision system. The key issue for vision system is proper segmentation and detection of an object under noise and low light. A number of image enhancement algorithms have been proposed so far such as adaptive histogram equalization (AHE) and contrast limit adaptive histogram equalization (CLAHE) to combat the low light and high noise in the captured image. However, both of these algorithms have some advantages and disadvantages based on image quality. Therefore to incorporate the advantages of both algorithms, we design a vision system actualization (VSA) system based on automatic image tuning (AIT) system that selects the suitable algorithm for image enhancement on the basis of image noise variance and light intensity. The VSA system is implemented on a 3-DOF robotic system using embedded control system (ECS). The proposed method shows optimum results up to 92 percent and minimizes the processing times up to 0.005 seconds for the VSA system. In addition, we developed a human-robot interface (HRI) to control and optimize the vision performance.