Lightweight Remote Sensing Target Detection Method Based on Improved YOLOv5
This paper proposes an improved lightweight remote sensing image target detection method based on YOLOv5. At the feature extraction end,depthwise separable convolution is used to replace the conventional convolution kernel to reduce the amount of compu-tation,and a coordinate attention mechanism is introduced to improve the learning strength of the model for small object features. At the feature enhancement end,a multi-layer fusion feature pyramid is used to replace the path aggregation network to obtain feature maps with more information richness. The experimental results show that the improved model proposed in this paper is significantly better than the original model and the comparison model in terms of detection accuracy,and at the same time,it shows good robustness to targets in different environments,and the amount of computational parameters generated in the test hardware environment is small,which can reach the level of real-time detection and can be deployed on edge computing hardware with limited computing power to car-ry out real-time inference detection.
万方数据
维普
