Railway Foreign Object Detection Using Feature Maps and Motion Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional railway foreign object detection methods are unreliable due to their sensitivity to light and camera angles, making them ineffective in complex outdoor scenarios, and they fail to provide timely warnings for sudden and unpredictable foreign object intrusions, which can lead to severe consequences.
Innovation Solution
A method and system utilizing a siamese network and UNet segmentation network to construct a foreign object intrusion prediction model, combined with a moving object filter model, to accurately identify and filter moving objects from image frames, enhancing the reliability of detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If traditional image detection methods are used to compare captured images with template images in real time, then the detection process is simple to implement, but the reliability is poor due to sensitivity to light and camera angles
Solution Approach 1:
The patent transforms the detection approach from direct image comparison to feature extraction and parameter comparison. By converting images into feature vectors and comparing key parameters (such as object presence, position, and characteristics), the system achieves reliability independent of lighting and camera angle variations while maintaining implementation feasibility through standardized feature extraction pipelines
2Device complexity
If traditional image detection methods are used, then the system structure is simple, but the adaptability to complex outdoor scenarios is poor
Solution Approach 1:
The patent segments the detection process into distinct modules: feature extraction, feature comparison, and result determination. This modular segmentation allows the system to adapt to complex outdoor scenarios by independently optimizing each module while maintaining overall system manageability and reducing complexity through clear separation of concerns
Solution Approach 2:
The patent introduces feature vectors as an intermediary representation between the captured image and the detection result. This intermediary layer abstracts the image data into scenario-invariant features, enabling the system to adapt to various outdoor scenarios without requiring complex scenario-specific processing while keeping the overall system structure relatively simple
3Ease of operation
If manual detection is used for foreign object intrusion, then the system is easy to operate, but timely warning cannot be given for sudden foreign object intrusion events
Solution Approach 1:
The patent replaces manual detection operations with automated computer-based image processing and feature comparison systems. This substitution eliminates the time delay inherent in manual detection while maintaining ease of operation through automated workflows that require minimal human intervention, providing timely warnings for sudden foreign object intrusion events
Data Source
AI summary
A method and a system for detecting railway foreign object intrusion, and a device and a medium can include acquiring a first key frame of railway image of a current scenario as a template image, and an Nth frame of railway image as a to-be-detected image; inputting the template image and the to-be-detected image of the current scenario into a foreign object intrusion prediction model to obtain a foreign object feature map of the current scenario, so as to determine whether there is a different target between the to-be-detected image and the template image of the current scenario; and filtering, by using a moving object filter model, a moving object from the different target of the to-be-detected image of the current scenario in response to existing different target, to obtain a true foreign object in the to-be-detected image of the current scenario.