Research Of Environment Perception System Of Mining Excavator Based On Binocular Vision

Mining excavator is an important equipment in open-pit mining,but due to the harsh environment on the mining site,the health and safety of personnel can not be guaranteed,skilled operators of mining equipment are difficult to cultivate,and the production efficiency is difficult to maximize,so the unmanned mining excavator attracts more and more attention.Three-dimensional perception of the surrounding environment is the key technology to realize unmanned excavator for mines.Relevant research on environment perception of excavator is of great significance to realize unmanned excavator.This paper studies the environmental perception of binocular camera used in excavator,and obtains three-dimensional data of surrounding environment by using the stereo imaging principle of binocular camera.In order to study the effect of binocular camera in mining excavator environment perception,this paper chooses the common ore heaps in mining scene as the experimental object,and establishes an environment perception experimental platform.The software components of binocular vision of perception system are determined,including interactive control module,image data acquisition module,camera calibration module,image preprocessing module,image correction module,stereo matching module,three-dimensional reconstruction module.The system software is mainly studied.In order to get the conversion relationship between the binocular camera image and the target being photographed,the binocular camera is calibrated.According to the characteristics of common camera calibration methods,Zhang Zhengyou camera calibration method is selected as the calibration method of binocular camera.The internal parameter matrix,distortion coefficient and external parameters of the left and right cameras are obtained by separate calibration of the left and right monocular cameras and binocular cameras.The calibration parameters are used to correct the distortion of the image and reduce the effect of image distortion on the three-dimensional reconstruction.Select appropriate grayscale and filter algorithm for image preprocessing,and stereo correction of binocular image to improve image processing speed and accuracy.A stereo matching algorithm based on SURF feature detection is designed.Stereo matching experiments are performed using binocular images of the experimental objects.The results are compared with those of FAST and SIFT feature detection algorithms.According to the principle of three-dimensional reconstruction,a three-dimensional reconstruction method based on parallax value is selected.The parallax map is used for three-dimensional reconstruction of experimental objects.Three-dimensional reconstruction experiments under different conditions were designed,and the effects of different illumination conditions and shooting distance on three-dimensional data were studied.In this paper,the scheme design of mining excavator environmental awareness system has been completed,and the related experimental studies have been completed.Experiments show that the stereo matching algorithm based on SURF has a good matching speed and effect in mining excavator working scene.Using binocular camera with even enough light and reasonable shooting distance can effectively improve the accuracy of three-dimensional reconstruction data.Binocular camera is used for three-dimensional reconstruction to obtain spatial three-dimensional point cloud data,and the three-dimensional point cloud images of experimental objects are generated,which achieves the research purpose of three-dimensional environmental perception of Excavators Based on binocular vision.