基于视觉图像的飞行区大型活动目标监视方法
By combining feature point extraction and Euclidean clustering algorithms with network models based on MobileNetV3 and YOLOv5, the problem of inaccuracy in large target surveillance in the flight area was solved, achieving stable and accurate target surveillance and contour estimation, thus improving airport security.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CIVIL AVIATION UNIV OF CHINA
- Filing Date
- 2023-09-21
- Publication Date
- 2026-07-17
AI Technical Summary
Existing target surveillance technologies cannot effectively handle large scenes, small targets, and situations with heavy obstruction within the flight zone. Furthermore, the positioning is based on point sources and cannot reflect the target outline information, resulting in inaccurate surveillance, easy jumps, easy loss of points, and difficulty in adapting to complex traffic scenarios.
We employ network models based on MobileNetV3 and YOLOv5 for target detection, combining feature point extraction and Euclidean clustering algorithms to estimate the contours of large targets. Contour segmentation is performed through feature point extraction and Euclidean clustering, and the target contour information is reflected through contour estimation.
It has improved the stability and accuracy of target surveillance in the flight area, reduced blind spots, provided clear and intuitive monitoring of target motion, made up for the shortcomings of point source positioning, and improved the accuracy and safety of the surveillance system.
Smart Images
Figure CN117475195B_ABST