Method and system for detecting hydraulic support propulsion based on monocular vision automatic labeling

Abstract

The invention provides a method and system for detecting the propulsion degree of a hydraulic support based on monocular vision automatic labeling. The method includes: installing a camera on the hydraulic support, and the camera shoots down the push rod of the hydraulic support, and the hydraulic support at both ends of the push rod Marker 1 and Marker 2 are respectively set on the base and the crosshead of the scraper conveyor; a ten-layer deep residual network model composed of one-dimensional convolutional layer Conv1D(M, N) is constructed; the deep neural network model is placed in the controller In the GPU unit, the controller regularly reads the image information of the camera, the position information of the camera and the distance of the push rod and sends them to the deep neural network model to train the deep neural network model; the controller reads the image information of the camera and the distance of the camera The position information is sent to the trained deep neural network model as an input signal, and the distance of the push rod can be output. The method of the invention does not need to be calibrated, does not need to provide labels manually, and can be trained online.

Description

technical field [0001] The invention relates to a detection method based on machine vision, in particular to a method and system for detecting the propulsion degree of a hydraulic support based on monocular vision automatic labeling. Background technique [0002] The hydraulic support is the core equipment for the safety support of the coal mine underground working face. In the working face, the hydraulic support will push the scraper conveyor forward after the shearer cuts the coal wall, thereby pushing the working face to the mining direction. [0003] During the mining process of the working face, the straightness of the scraper conveyor directly determines the thickness of the coal wall cut by the shearer drum, and then determines the coal mining volume of the single knife of the coal cutter. Since the straightness of the scraper conveyor is determined by the push cylinder of the hydraulic support, its straightness determines the spatial distribution of the hydraulic su...

AI Technical Summary

Problems solved by technology

However, the monocular camera can move in two dimensions, the horizontal axis and the pitch axis, so the calibration method of the monocular camera cannot be used to determine the imaging distortion under dynamic different angles

In addition, during the propulsion process of the hydraulic support, the position of the monocular camera relative to the ground will also change, resulting in a large difference in the shooting environment of the camera

Therefore, using a monocular camera, the accurate detection of the stroke of the hydraulic support cannot be achieved through the calibration method

Images