Research On Auti-collision Warning System For The Mine Electric Vehicle Based On Monocular Vision

Electric vehicle is the main transportation in coal mine,its security issue has raised more and more attention. Therefore, this paper designs an anti-collision warning system for electric vehicle, using monocular camera and machine vision algorithms which are suitable for the mine roadway environment. The system achieves the identification of the obstacles in front of the electric vehicle and distance measurement. By setting some anti-collision measures, the system can realize early warning. The main work of this paper includes:a. Obstacle image preprocessing. Firstly,considering the characteristics of the mine roadway image,this paper makes a noise smoothing about the image from camera, and adopts Retinex algorithm based on wavelet decomposition to enhance the image contrast. Then the obstacle threshold segmentation is made by using the Otsu combined with genetic algorithm. Finally,the general outline of obstacle is extracted by edge detection and morphological processing.b. Features extraction and identification classification of obstacles. Firstly, the geometry shape parameters are used to locate the obstacle accurately and remove irrelevant interfere of the background. Then this paper proposes the improved moment invariant algorithm to extract the invariants moment of obstacle. Finally, the geometry shape parameters and Euclidean distance similarity are adopted, to realize the classification and recognition of obstacles.c. Distance measurement of obstacles. Firstly, the imaging model of the camera is analyzed. Then the camera is calibrated, and the camera’s internal parameters are obtained. Finally the model of distance measurement is built according to the principle of pinhole imaging, to calculate the distance between the electric vehicle and the obstacles.d. Design of auti-collision warning system. The system software is designed by adopting modularization idea, according to the requirements of electric vehicle driving speed, safe distance and program running time.