Abstract: In the process of condition monitoring and fault diagnosis of substation equipment,the infrared thermal image recognition technology that can automatically identify substation equipment is one of the key technologies.In order to solve the problems of excessive concentration of background temperature,low contrast and lack of intelligent methods in the current infrared thermal image recognition of substation equipment,a method of image enhancement using RetinexNet algorithm is proposed to create conditions for accurate recognition of infrared thermal images. Object detection is performed on the enhanced image using YOLOX-darknet53 algorithm. In the experiment,the infrared thermal images are recognized by used the proposed method. Not only the recognition time can reach 6.88 ms each piece of image, but also the average accuracy of 8 kinds of substation equipment can reach 96.51%.The experimental data show that the method is efficient and accurate,and can meet the needs of monitoring the status of substation equipment.