A coal gangue recognition method based on high-speed computed spectral imaging

By constructing a coal gangue identification method based on high-speed computational spectral imaging, and using a fixed diffuse reflection reference bar and discrimination axis for drift compensation, the problem of identification instability caused by light source power fluctuations and imaging chain changes is solved, thereby improving the stability and accuracy of coal gangue identification.

CN122391992APending Publication Date: 2026-07-14HENAN ACAD OF SCI INST OF APPLIED PHYSICS CO LTD +2
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HENAN ACAD OF SCI INST OF APPLIED PHYSICS CO LTD
Filing Date
2026-04-22
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

In existing technologies, fluctuations in light source power and changes in the imaging chain during coal washing and dry separation processes lead to unstable coal gangue identification results, making it difficult to balance identification accuracy and engineering applicability under high-speed conditions.

Method used

A method based on high-speed computational spectral imaging is adopted. By obtaining the dimensionless coded response vector of a fixed diffuse reflection reference bar, a linear inversion matrix and a discrimination axis are constructed. Drift compensation and stability discrimination are performed, and the coal gangue identification result or conservative sorting result is output.

Benefits of technology

It improves the stability and accuracy of coal gangue identification, suppresses the interference of spectral response drift on sorting and judgment, and enhances the identification effect under high-speed operating conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122391992A_ABST
    Figure CN122391992A_ABST
Patent Text Reader

Abstract

The application relates to the technical field of coal gangue recognition and intelligent sorting, and discloses a coal gangue recognition method based on high-speed computing spectral imaging. In the coal gangue recognition process of high-speed computing spectral imaging, the recognition result of coal and the recognition result of gangue are affected by slow drift of an imaging chain, and the boundary of the two recognition results is unstable. A progressive processing chain composed of a fixed diffuse reflection reference bar, a coal-gangue discrimination axis, a normalized discrimination axis, a running period discrimination drift coefficient and a drift constraint recognition score is constructed. The running period discrimination drift coefficient is extracted in a unified discrimination description space, and the target discrimination description vector is compensated for drift along the normalized discrimination axis according to the running period discrimination drift coefficient. In combination with a drift effectiveness marker, a coal recognition result, a gangue recognition result or a conservative sorting result is output. The method can effectively inhibit the interference of illumination fluctuation, channel drift and local observation anomaly on sorting determination, and improve the stability, accuracy and engineering applicability of coal-gangue recognition under high-speed running conditions.
Need to check novelty before this filing date? Find Prior Art