UAV aerial photography and imaging technology combines the aerial superiority of UAV and the intelligent analysis ability of computer vision, which can be widely used in geographic information, agriculture, environmental protection, infrastructure monitoring and other fields, providing efficient and accurate data support for all walks of life, which has important applications and significance. However, there are problems such as the size and accuracy of target recognition cannot be taken into account, and there are problems such as large difference between target detection object recognition and actual size, and there are problems such as occlusion. In order to improve the accuracy of UAV aerial photography in real-time vehicle monitoring, this paper inserts an efficient Vehicle Detection Aggregation Network (ELAN-VD) into its backbone layer network based on the YOLOv7-tiny model, so as to achieve real-time monitoring of vehicles on the road with low cost, high recognition speed and high recognition accuracy. Experiments show that the improved model mAP can be as high as 58%, which is much higher than that of the YOLOv7-tiny model, and the low cost of this model further illustrates its practical application value.
Research on High-Precision Detection of Road Vehicles in UAV Aerial Photography Based on YOLOv7
27.09.2024
1172825 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
MS-YOLOv7:YOLOv7 Based on Multi-Scale for Object Detection on UAV Aerial Photography
DOAJ | 2023
|Improved YOLOv7 Target Detection Algorithm Based on UAV Aerial Photography
DOAJ | 2024
|Road pothole detection based on improved YOLOv7
SPIE | 2023
|Automatic aerial photography system of high-precision aerial image
Europäisches Patentamt | 2018
|