Research Of Building Crack Monitoring Based On Digital Image Processing Technology

Building surface crack is common on the building structure. Its emergence affectsthe usability of structure, damages the wholeness of structure, or reduces the service lifeand may even cause safety accidents, making people’s life and property losses,adversing social impact. Therefore, building crack width monitoring is an importanttask in building health inspection.But for a long time, fracture monitoring of building health inspection is mostlytraditional methods, which are dangerous, time-consuming, and inefficient. So it isnecessary to develop a certain automated functional construction cracks non-contactmonitoring method.Due to preciously research on digital image processing techniques of building crackmeasurement method, and cracks in the complicated surroundings, it directly on thetreatment of fracture way may lead to the instability of the measured results. This paperwill research crack measurement method based on digital image processing techniquesfor further study, adopt a logo calibration to calibrat crack, and complete its finalrealization based on digital image processing technique to measure the building crackmethod, after the study of its realization method.This paper completed the following work:①This measurement principle is discussed, by monitoring observation marks therelative displacement of the crack width and its variation. It is studied theimplementation process, to get the measurement method of the specific implementationsteps, which is to set up in preparation for marks, cameras, lighting, observation area.Then it carries on the image acquisition, cuts image in image processing, imagepreprocessing, image denoising, angular point, pixel calibration. At the end of the paperis calculating crack width.②In this paper, the method of measurement is the important step in fractureangular point of observation symbol extraction in image. The common three kind ofcorner extraction algorithm (Moravec, Harris, SUSAN corner detection operator)principle were expounded, and its implementation program is realized by MATlab,allowing them to solving real shooting environment observation symbol of angularpoint. Based on their accuracy, precision, stability, complexity, Harris corner so as to bepickded out as the measuring method of extraction of operator. ③In order to understand the measuring accuracy of this measurement method, thispaper analyzes the observation symbol representation, pixel calibration precision andthe camera lens distortion mechanism of the effect of this method ultimatelymeasurement accuracy, and thus puts forward a method to reduce the impact. In thesame time, this paper calculates the part of the size of the influence factors.④This paper completes the software of this measuring method. After thismeasurement software structure design, it uses MATlab GUI to finish in the end of themeasuring methods of implementation.