Design of intelligent inspection robot system for electrical equipment based on visual recognition

Video surveillance has long been introduced into the remote monitoring of the operation state of power equipment. The early image monitoring in the production environment realized by “remote viewing” and robot only realized the digitization and long-distance transmission of image and sound signals, which provided comprehensive technical support for solving the remote multimedia monitoring system in the production environment. However, it lacks the function of information processing, and can not intelligently and actively identify and analyze the monitoring targets. The operator needs to observe and analyze the images at any time, which virtually increases the burden of the dispatcher. In addition, manual judgment is subjective and human eyes are prone to fatigue, which is difficult to meet the requirements of automatic monitoring of power equipment operation status. To a certain extent, it ignores the accurate judgment of accident development afterwards and reduces the rapid response ability to accident treatment. The time for eliminating defects is delayed, affecting the power sales and service commitment.