Particle distribution detection method and device based on particle density
By estimating particle density and adaptively adjusting cutting parameters through a deep learning image segmentation model, combined with global coordinate system transformation and deduplication algorithms, the problem of balancing particle detection accuracy and efficiency in existing technologies is solved, and efficient and accurate particle distribution detection is achieved.
Patent Information
- Application Number
- CN202510721987.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-30
- Publication Date
- 2025-09-12
AI Technical Summary
When processing large-size, high-density particle images, existing technologies have difficulty balancing detection accuracy and computational efficiency. The fixed overlapping cutting method leads to a high particle missed detection rate, serious waste of computing resources, and insufficient adaptability.
Particle density is estimated through a pre-trained deep learning image segmentation model, image segmentation parameters are adaptively adjusted, and the image segmentation method is optimized by combining global coordinate system transformation and deduplication algorithms to reduce missed particle detections and improve detection efficiency.
It realizes the dynamic adjustment of cutting parameters according to particle density, improves the accuracy and efficiency of particle distribution detection, and is suitable for real-time processing of high-resolution images.