Coal flow debris rotation detection method based on physical fuzzy modeling and scene customization attention

By combining physical fuzzy modeling with scene-customized attention-enhanced feature extraction, the accuracy and robustness issues of existing coal flow debris detection under high-speed motion fuzzy conditions are solved, achieving high-precision debris detection and positioning, which is suitable for intelligent sorting systems under complex coal flow conditions.

CN122312993APending Publication Date: 2026-06-30CHINA UNIV OF MINING & TECH
0 Cites 0 Cited by

Patent Information

Application Number
CN202610509781.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-17
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

Existing coal flow debris detection technologies lack anti-interference capabilities under high-speed motion ambiguity conditions, making it difficult to effectively extract weak features and accurately locate rotating targets. This results in missed detections, false detections, and low positioning accuracy, failing to meet the high-precision requirements under complex working conditions.

Method used

A detection method based on physical fuzzy modeling and scene-customized attention is adopted. By constructing a motion fuzzy mapping model and a scene-customized attention module, feature extraction and background interference suppression are enhanced. Combined with rotating bounding boxes and non-maximum suppression logic, the final detection result is output.

Benefits of technology

It improves the robustness and accuracy of detection in complex backgrounds, reduces the false negative and false positive rates, and outputs accurate detection results including category, confidence level, rotated bounding box and actual physical size, supporting subsequent sorting and gripping operations.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention provides a coal flow debris rotation detection method based on physical fuzzy modeling and scene customization attention. The method comprises the following steps: acquiring a clear coal flow image, and labeling a non-coal debris target in the clear coal flow image; establishing a data set; constructing a sundry detection network; constructing a physically-driven motion fuzzy mapping model; performing dynamic fuzzy processing on the clear coal flow image by utilizing the motion fuzzy mapping model, and generating a training sample for simulating the working condition of the motion fuzzy phenomenon generated by the conveyor belt; training and optimizing the sundry detection network; inputting a real-time coal flow image into the optimized sundry detection network to obtain a candidate detection result of a non-coal sundry target; and post-processing the candidate detection result and outputting a final screening detection result. According to the method, the stability and accuracy of impurity detection under the complex working condition of high-speed coal flow can be improved, the influence of motion blur and background interference is reduced, and technical support is provided for safety monitoring and automatic sorting of a coal mine conveying system.
Need to check novelty before this filing date? Find Prior Art