Technology for measuring mineral particle size based on multi-element multi-scale entropy

Abstract

The invention discloses a technical method for measuring a mineral particle size based on multi-element multi-scale entropy. The method comprises the following steps: firstly, respectively utilizing a sobel operator to enhance boundary, for three color variables R, G and B of coarse crushing, intermediate crushing and fine crushing mineral images; averaging each row of gradient data of ternary color variable of each level image; continuously intercepting an average row sequence of each color variable in fixed length according to the sequence; respectively averaging the data in each particle size data set of the ternary color variable; constructing a (n-m) set ternary compound delaying vector at the scale E and calculating a multivariate sample entropy, thereby acquiring multiple multi-scale entropy at 20 scales; lastly, establishing an associated model of the entropy and the mineral particle size utilizing the calculated multivariate sample entropy value of each-level mineral image gradient data, thereby detecting the mineral particle size under each-level crusher.

Description

technical field [0001] The invention relates to particle size detection in the industrial field, especially in the field of mining and metallurgy, and relates to a technology for particle size detection using multivariate and multi-scale entropy in detection images. Background technique [0002] With the development of mineral processing technology, the requirements for ore technology are becoming more and more complex. The crushing process is generally to pass the excavated ore through a multi-stage crusher to detect the particle size of the ore, and then transmit the ore with a qualified particle size to the downstream for subsequent processing. Crushing includes multi-stage crushing such as coarse crushing, medium crushing, and fine crushing. If the particle size distribution of the crushed products is unreasonable during the crushing process, problems such as full or empty bins of the crusher will occur, resulting in the failure of the crusher to operate normally. Howev...

AI Technical Summary

Problems solved by technology

Large dynamic range, fast measurement speed, good repeatability, and convenient operation, but the resolution is low, and the accuracy is poor when measuring samples with a narrow particle size distribution range

(3) The particle size detection method based on image processing, using digital image processing methods to segment and identify ore images, due to the overlapping and accumulation of ore, there are many ore adhesions in the segmented image, and because in the image acquisition process, The resolution of the camera is difficult to identify the situation where the small ore is stuck together, and the small ore accumulation area is removed to avoid affecting the subsequent processing, but the small ore accumulation area can still be segmented at a higher resolution, resulting in a certain impact on the detection accuracy

Images