Digital Image Recognition For Electrical Control Cabinet Switch State Detection

With the development of the power system, power equipment is moreand more complex, and the correlation of the various subsystems is becomingmore closely. Supervision to subsystems is increasingly important. So moreand more requirements are need in automation field. Based on such a demand,digital image recognition for electrical control cabinet switch state detection iscreated and developed.The digital image recognition technology is an important class oftechnology in modern IT technology. It is a hot research topic in theinformation systems discipline. This paper studies the feasibility of usingimage detection in electrical switchgear, using the application of digital imagerecognition methods, such as image preprocessing (gray histogram smoothspatial filtering, sharpening spatial filtering), morphological detection andlocation and edge detection to detect the state of the switch of the electricalswitch cabinet automatically and determine whether the switch state is defaultstate.This paper analyzes the basic theory of digital image and graphicsprocessing algorithms. Based on the switching characteristics of intelligentcontrol the cabinet switch state identify trends, exploration of positioningtechnology and state characteristics, we propose the switch identificationsystem. Shadow removal system image preprocessing, color modelbinarization, and switch positioning, switch state recognition comparison areused in this paper. By the experimental results, the system can achieve anaccurate positioning of the switch, the correction and segmentation andrecognition.