Based On Fractal Edge Of Pavement Surface Distress Image Recognition Research

Currently the conservation and management of domestic and international road andmainly based on manual inspection from automatic detection of data to implement. So athome and abroad are investing huge costs for research and development of all kinds of roadautomatic acquisition equipment and data processing system, are generally appliedmulti-function road check-out car system for pavement damaged image, rut, smoothness, andthe surrounding environment information, subsequent analysis of data processing andrecognition, so as to provide maintenance and repair report. Pavement damage informationprocessing of the data due to the complexity of the road itself and in the process of collectingdata of the external environment, makes the pavement damage image data processingworkload is huge and very long time. So on roads damaged image automatic identificationsystem research is of great significance.The content of this research is for equipment acquisition of pavement damage imageautomatic recognition, mainly for all kinds of cracks in pavement damage images forautomatic identification. To identify the crack in the images of the damaged well information,this paper puts forward two kinds of fractal method of pavement damage in the crack in imageinformation recognition, is a fractal self-similarity of pavement damage image recognition,another is based on the fractal theory in the discrete fractional Brownian motion model foridentification. Studied how the damaged image of logo marking and rutting information suchas the removal as well as to the damaged road surface crack information in imageenhancement and image segmentation, finally achieved the road surface crack in the imagesof the damage identification research purpose.In this thesis, the results of fractal models can be created with a good description of thetexture mage; The image acquisition surface is pretreated in order to achieve the purpose ofimproving image quality, the final crack damage the image edge detection and automaticidentification of the road to prepare; For roads damaged image and self-similarity of thecharacteristics of self-affine, based on fractal theory damaged pavement crack edge imagerecognition and detection. According to road damage fractal image features proposed crackidentification DFBR model self-similarity and fractal image based on the damaged road,which identify the damaged pavement crack image information.