Study on Image Registration and Image Matching of the Robot System for Substation Equipment Inspection
|School||Harbin Institute of Technology|
|Course||Computer Science and Technology|
|Keywords||Image Registration Matching based on feature points Chamfer Matching|
In the mobile robot system for substation equipment inspection, the computer analysis and processing for images shot by robot is the core of all the components. Through the state recognition for images of equipments, it can automatically monitor the working conditions of a substation. For the same scene, there are much correlation and certain difference between the images shot by robot. So, to find the location of equipment, that is to determine the location of target become the key to the study. Image registration can effectively solve this problem.Image registration is a basic problem in image processing. It matches two or more images, at different times, from different sensors or from different view points, in order to remove of the potential geometrical distortion existed. Image registration is extensively applied to computer vision, pattern recognition, medical image processing and remote sensing information processing.This article researches on the image registration based on characteristic points for the mobile robot system for substation equipment inspection. In response to needs of the problems, we discuss a method based on multi-bound matching which takes Harris corner points as feature points, and uses the grey level grows matrix’s Bhattacharyya distance as the similarity. On the basis of this, RANSAC matching is to be done. This method has effectively improved the accuracy and speed of whole process of matching.This paper presents a simple and effective method of image matching. The method uses the result of image registration as a initial assumption of the image Chamfer matching, using the original Chamfer edge matching method, calculating the similarity between the images, to achieve the goal to measure the results of image registration. Experiment proved that in selecting appropriate threshold, there are very satisfactory results.In this paper, we use image registration to determine the location of target, and use image matching to measure the results. According to the results, we make the appropriate treatment of the images of equipments. We design the image processing and recognition algorithm of the switch, switches, meters, and other substation equipment, and to achieve automatic image processing and pattern recognition function in the mobile robot system for substation equipment inspection.