Laser radar and camera fused artificial intelligence ore identification method and device

Abstract

The invention discloses a laser radar and camera fused artificial intelligence ore identification method and device, and the method and device are higher in applicability and are good in real-time performance. According to the specific scheme, ore is placed on a conveying belt running at a constant speed; a two-dimensional laser radar is adopted to emit a laser signal to the ore, and point cloud data information of the ore is acquired according to an ore return signal, wherein the point cloud data information comprises a distance point cloud and a reflectivity point cloud. A linear array camera is matched with an illumination light source to collect camera images of ores. Pixel-level fusion is performed on the point cloud data information of the ore and the camera image to obtain a three-channel image of the ore, wherein the three channels are respectively a gray value channel, a distance point cloud channel and a reflectivity point cloud channel. The three-channel image of the ore isnormalized, eliminating an interference background is eliminated, denoising processing is performed, texture features, three-dimensional shape features and gray features are extracted, and the extracted features are identified by using a convolutional neural network to obtain an ore identification result.

Description

technical field [0001] The invention relates to the technical fields of photoelectric detection, photoelectric imaging and photoelectric measurement, in particular to a laser radar and camera fusion artificial intelligence ore identification method and device. Background technique [0002] In terms of ore mining, my country's annual ore output is huge, and the demand for sorting ore / miscellaneous materials and coal / gangue is also huge. According to statistics, by the end of 2018, the output of coal alone will reach 3.97 billion tons per year. Taking the coal industry as an example, the current coal mining has changed from traditional manual mining to modern semi-automatic mechanical mining. The disadvantage of high intensity and the harsh sorting environment pose a great threat to the health of the sorting personnel. [0003] At present, the imaging identification methods in the ore field mainly include high-definition camera imaging, infrared camera imaging and X-ray imag...

AI Technical Summary

Problems solved by technology

Therefore, the existing ore identification methods have poor applicability and poor real-time performance

Images