Target object detection method and device
By acquiring video frame images and radar data on the train, extracting and fusing the track area to identify target objects, the problem of high computational complexity in multi-sensor detection is solved, and real-time and accurate obstacle detection is achieved.
CN120808305APending Publication Date: 2025-10-17SHANGHAI RENTONG ELECTRONICS TECH CO LTD
0 Cites 0 Cited by
Patent Information
- Application Number
- CN202510827892.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-19
- Publication Date
- 2025-10-17
AI Technical Summary
Technical Problem
In the existing technology, when multiple sensors are used for obstacle detection, the amount of data is huge and cannot meet the requirements of real-time and accurate detection.
Method used
By acquiring video frame image sequences and radar data, the short-range and long-range track areas are extracted, the region of interest is set, and the radar data is fused with it to identify the target object.
Benefits of technology
It realizes real-time and accurate identification of obstacles and pedestrians on the train track, reduces the amount of calculation, and improves the real-time and accuracy of detection.
✦ Generated by Eureka AI based on patent content.
Abstract
The invention discloses a target object detection method and device. The method belongs to the technical field of obstacle detection, and comprises the following steps: acquiring a video frame image sequence and radar data collected by a train in the running process of a running track, the video frame image sequence comprising at least one video frame image; extracting a short-distance track area and a long-distance track area in the video frame image; fusing the short-distance track region and the long-distance track region to obtain a region of interest; the interested area and the radar data are fused to obtain interested data, and the interested data comprise two-dimensional data corresponding to the video frame image and three-dimensional data corresponding to the radar data; and identifying a target object in the region of interest from the data of interest. According to the embodiment of the invention, the target object on the traveling route is accurately detected in real time in different scenes, and the possibility of danger occurrence is reduced.
Need to check novelty before this filing date? Find Prior Art