Research of Detection Method of Mine Motor Vehicle Pedestrian Based on Image Processing

摘要

This paper use infrared camera to collect the front image of motor cars, and to pretreat image based on genetic algorithm and normalized incomplete Beta function. Using pulse coupled neural network for image two value segmentation; and using improved fuzzy edge detection algorithm based on genetic algorithm for recognition the rail and using heuristic method for fitting the rails; once pedestrian recognition algorithm identified pedestrian, the alarm is immediately triggered. This system can efficiently identify the pedestrian near the track, judge and early warn the position of pedestrian; it is a new technology which can eliminate the potential safety hazard of motor vehicles in the transportational process.