Substation instrument detection with improved YOLOv5 algorithm based on AI front-end

With the progress of robot technology, the application of intelligent robot in substation inspection is more and more common. However, the current target detection algorithms have too many parameters, which makes it difficult to realize real-time detection on embedded platforms. Therefore, an improved YOLOv5 substation instrument detection algorithm based on AI front-end is proposed. This algorithm introduces CBAM fusion attention mechanism module, which can effectively improve the ability of the model to extract important features, thus improving the accuracy of detection. The algorithm is reconstructed and optimized by TensorRT technology to reduce the computation and increase the detection speed. Experimental results show that the algorithm achieves a detection speed of 24FPS on the NVIDIA Jetson Nano platform with slightly reduced accuracy, which provides a possibility for real-time instrument detection in edge computing environment.