ISSN 1000-0526
CN 11-2282/P
CN 11-2282/P
Identification and tracking of bird clutter in weather radar data based on YOLOv5 and DeepSort
- Article
- Figures
- Metrics
- Preview PDF
- Reference
- Related
- Cited by
- Materials
Abstract:
According to the specific image feature that the bird echo shows obvious ring shape on the Weather Radar reflectivity product, an improved algorithm based on a lightweight convolutional neural network You Only Look Once Version5（YOLOv5） and multi-object tracking based on Deep learning based Simple Online and Realtime Tracking（DeepSort） is proposed to identify ,the?training?and?test?datasets?are?constructed?fromradar?volumetric?scanning?echo?intensity data?obtained?from?the?Yingkou Weather?Radar?from?2020?to?2023, track the bird echo respectively. Firstly, Shuffle Attention（SA）, a lightweight attention mechanism, is introduced into YOLOv5 algorithm to improve the accuracy and effectiveness of the overall model detection. Secondly, in DeepSort algorithm, the original cross merge ratio Intersection over Union（IOU） matching mechanism is replaced by an improved the loss function of object detection Distance-Intersection over Union（DIOU） matching mechanism. DIoU introduces the distance between the center points of the boundary box on the basis of calculating the overlap degree of the boundary box, so as to provide more accurate positioning. The number of identification （ID） error matching and ID switching caused by partial occlusion overlap is reduced. The experimental results show that the optimized YOLOv5 algorithm improves the accuracy by 2.6%, the recall rate by 1%, and the average accuracy of threshold values greater than 0.5 by 1.2%. The improved DeepSort algorithm reduces the number of ID switches by 2 times, multi target tracking accuracy Multi-Object Tracking Accuracy（MOTA） increases by 4.5%, the improved lightweight of the initial model, and the overall detection performance is significantly improved, meeting the actual demand for bird echo recognition and tracking.
Keywords:
Project Supported:
Clc Number:
Mobile website
