Font Size: a A A

Technology Research Of The On-line Detection Of The Length Of The Vehicle Based On Sequence Image

Posted on:2015-03-10Degree:MasterType:Thesis
Country:ChinaCandidate:F FengFull Text:PDF
GTID:2348330485491785Subject:Instrument Science and Technology
Abstract/Summary:PDF Full Text Request
In recent years, overload and illegal modification of vehicle result in the traffic accidents and urban traffic congestion, which have bad influence on traffic. In this paper, we further stuy the domestic and foreign technological advances of the dimensions of the vehicle and explore the surrounding of the measurements, with considering the performance requirement. This paper researchs design project of the measuring system of the dimensions of vehicles and completes the establishement of the measuring system. We present a method of the on-line detection of the length of the vehicle based on image sequences. The main research contents include:1. Considering the factors such as development costs, hardware structure, measuring accuracy, response speed and antijamming capability, this paper presents a method of the on-line detection of the length of the vehicle based on image sequences, with machine vision and infrared laser detection. It analyzes the factors, which affect the measuring accuracy, such as angle of field of view, frame rate, the error of image mosaics, angle correction.2. This paper researchs the image processing algorithm of edge detection and feature points extraction based on vehicle's image sequences. It uses Sobel filter to detect the image edges and also uses Harries filter to extract the feature points of the images of the vehicle. The background difference is used to extract foreground image and remove the disturbance based on class variance. This algorithm also removes the noise of the vehicle's surface by connected component labeling.3. This paper researches the measuring theory and the mathematical model of the image matching based on pixel correlation of gray and the image mosaic based on weighting coefficient. We employ the a hierarchical search strategy to search the matching points, which can improve the efficience of the image stitching. We do the experiments, which rebuilds the full view of the vehicle's body, of the measurement of dimensions of the vehicle based on theory to verify the feasibility of the project.4. This paper researches the geometric model and the calibration method of the camera. According to the features of the measuring environment, we choice the multi-plane calibration method to conduct the camera calibration experiment and work out intrinsic parameters, external parameters of camera and lens distortion coeffcient. We conduct the experiment of the measurement of the dimensions of vehicles with different-sized, different types cars as the measuring objects. With data and statistics, the influence factors of accuracy of measurement of vehicles' length.
Keywords/Search Tags:Machine Vision, Image Sequence, Length of Vehicle, Edge Detection, Image Matching and Stitching
PDF Full Text Request
Related items