Multi-weather Route Detection System Of Overhead Tyre Crane

Currently,overhead tyre crane(hereinafter referred to as "Overhead tyre crane")is the important tool of container vertical operation in the whole terminal yard,the route of tyre crane mainly depends on the naked eye for judgment,which is easy to lead to misjudgment in complex weather,with its poor safety,low efficiency and weak competitiveness.By studying the application of visual navigation in lane line recognition,this paper proposes a corresponding road line detection algorithm for normal light,strong and weak light,rainy and haze days,and designs a route detection system of overhead tyre crane,which is of high timeliness,good stability and adaptability to a variety of weather conditions.The specific research contents are as follows:1.In order to improve the real-time performance and accuracy of lane line detection,this paper firstly compares and analyzes basic road line detection methods named interception of interest areas,image gray scale,direction adjustment and filtering enhancement,binarization,and edge feature detection.Secondly,for the images under different light conditions,it classifies images with different light intensity.It uses three-segment linear gray stretch method to enhance the contrast of images with strong light,and uses MSR algorithm to enhance the details of images with weak light.For road images in rainy days,it uses image dewatering algorithm based on guidance filtering.For road images in haze weather,it adopts dark channel prior algorithm for image defogging.When the pre-processing and enhancement of images in various weathers is completed,the lane line detection is performed.The experimental results show that the algorithms in this paper solve the problems of multi-weather route detection and provide a basis for subsequent lane line fitting.2、The problem of lane line fitting and tire crane attitude estimation is studied.In order to improve the speed and accuracy of lane line fitting,this paper proposes a lane line detection method of probability Hough transform under the constraint of polar angle.The mean and variance of lane line direction angle are obtained by statistics to determine the polar angle range of probability Hough transform.It verifies that the algorithms can effectively reduce the search range and improve the detection speed through experimental comparison.Tire crane attitude estimation uses the two parameters of lane line deviation angle and yaw value to determine the real-time status of the tire crane,deviation warning of the attitude of the tire care by setting a safety threshold,and theoretically and experimentally tested its rationality.3、Designed the upper computer interface of overhead tyre crane.By using the PyQt,the humanized upper computer interface is designed to realize detection control module and lane line information display module.The detection control module mainly realizes the selection function of processing road images in different lighting,rainy and foggy days,while the lane line information display module displays the lane line information filtered out and tire crane attitude parameters in real time.4、The experimental prototype model was built to simulate the working environment of the tyre crane,and the feasibility of the system is tested.The lane line identification experiment was carried out on different weather lines during the experiment The system can accurately and quickly identify lane lines in the situations of normal light,strong or weak light,rainy days and foggy days.The experimental results show that its real-time and robust performance of road image processing can meet the working requirements of overhead tyre crane.This project has important theoretical significance and application value for assisting driving of overhead tyre crane.