A vision-based method for detecting and controlling slurry flow in a spiral concentrator.
By constructing a slurry edge recognition model in a spiral concentrator, using image segmentation and Kalman filtering algorithms to identify the slurry edge position, and combining this with a geometric relationship model to calculate the flow rate, the problem of low flow control accuracy in existing technologies is solved, and automated and precise slurry flow rate adjustment is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JIANGXI UNIV OF SCI & TECH
- Filing Date
- 2025-12-30
- Publication Date
- 2026-05-26
AI Technical Summary
In existing technologies, the determination and adjustment of slurry flow rate in spiral concentrators rely on the operator's experience, resulting in large errors, low control precision, and an inability to achieve stable and accurate flow control.
By acquiring images of the slurry at the inlet of the spiral concentrator, a slurry edge recognition model is constructed. The edge position of the slurry is identified using an image segmentation algorithm based on UNet and a target tracking algorithm based on Kalman filtering. Combined with a geometric relationship model, the slurry flow rate is calculated to achieve automated flow control.
It achieves precise automatic control of slurry flow, improves the accuracy of judgment and adjustment, reduces reliance on human experience, and improves the efficiency and accuracy of the entire process.
Smart Images

Figure CN122089645A_ABST