Research On Texture Depth Algorithm Of Asphalt Pavament Based On Line Structured Light

With the vigorous development of the transportation, the highway mileage of China isbecoming more and more huge, frequent traffic accident makes the requirements of drivingsafety and comfortable also more and more high. At the same time, it also brought pressure tothe traffic administrative department, how to detect and maintenance the pavement quicklyhas become a key issue. The traditional detecting methods, such as sanding, penduluminstrument etc, cost much time and influenced by external conditions serious. It can’t meet thedemands of social development now. The application of technology about non-contactmeasurement based on laser and3d machine vision can effectively avoid the defects of thetraditional method, real-time and effective detect the pavement. It can reconstruct3d imageof pavement, makes the analysis convenient and visual.This paper first mainly discussed the processing technology of image obtained by the3dmeasurement system, introduced some core technology, such as pavement image analysis,image denoising and interference elimination. Then stated how to get texture depth by thedata after image processing.This paper first introduced the principle of optical3d measurement system, andsummarized the testing process of this system. Then focused on the analysis of the pavementimage processing technology, proposed a high-precision image thinning algorithm aftercontrast and comprehensive analysis. Main steps of the algorithm is as follows:1) Using fixedthreshold filter and the average threshold method to eliminate the influence of backgroundlight and low grey value of pixels;2) To eliminate the influence of noise by the way ofinterference elimination method;3)Completing image thinning by gravity method, and gettingthe structured light image centroid position.This article also expounded how to obtain the texture depth through the data after imageprocessing. This algorithm is based on the thought of sanding method, using integral methodto calculate the texture depth. First, cant eliminating reduce the influence of the cant causedby residue and the installation car during the detecting.Then, using sliding window and threshold to determine the space plane formed by the testdata. Finally, using integral thought to calculate the texture depth of the measurement area. Through test and data, this paper demonstrates the feasibility and effectiveness of the imagethinning algorithm and texture depth calculating algorithm. The algorithms can use the dataanalyzing of the texture depth detecting system and the system used in other fields.