Coal gangue photoelectric separation image particle aggregate area shape selection identification method

Abstract

The invention discloses a coal gangue photoelectric separation image particle aggregate area shape selection identification method. The method comprises the following steps: firstly, carrying out binarization processing on an obtained X-ray image; then extracting shape feature parameters of the binarized image, the convex hull binarized image, the concave defect binarized image and the edge image thereof; selecting the features by adopting a Relief algorithm to obtain an optimal shape selection feature combination; and finally, establishing a support vector machine classification model based on a linear kernel function. And the trained model is used for identifying an unknown classification aggregate area or an independent target area. The method not only can be used as a condition for judging whether the aggregate region needs to be further segmented into independent particle targets, but also can be used as a condition for judging whether the segmented aggregate region is the aggregate region.

Description

technical field [0001] The invention relates to the technical field of coal sorting, in particular to a method for identifying the shape of particle aggregate area in photoelectric sorting of coal gangue images. Background technique [0002] Based on dual-energy X-ray photoelectric separation technology, it is of epoch-making significance to upgrade coal gangue "group selection" to "single selection". The premise of identifying and separating dual-energy X-ray image targets is to obtain the region where the image targets are located, that is, to segment the target from the background, and to ensure that the segmented targets are all independent individuals. At present, for dual-energy X-ray coal gangue sorting systems, queuing devices are often used to distribute coal gangue to avoid sticking and overlapping. However, with the increase of the material input amount and the input size ratio, the randomly arranged coal and gangue will inevitably overlap and stick together. Th...

AI Technical Summary

Problems solved by technology

However, with the increase of the amount of material input and the proportion of input scale, it is inevitable that coal and gangue arranged randomly will overlap and stick together.

Images