Research Of Image Recognition Processing And Measuring Technology In Bridge Crack Detection

Concrete is the common building material of bridge. The evaluation of the durability of concrete is often associated with the existence and size of the crack. The presence of cracks on the bridge does great harm to the bridge, so crack detection is an important part of the bridge inspection in the world. Nowadays, the detection of bridge cracks mainly depend on manual work which is time-consuming、strenuous and dangerous. For the above mentioned reasons, based on image recognition processing and measuring technology, this paper proposes a method for underside crack detection, the author developed a software which is used for the recognition and measurement of the cracks at bottom of the bridge, these improved current related methodology.By using image processing techniques, author identified and measured the image and video which were collected by CCD (Charge-coupled Device) camera and image capture card at bottom of bridge. With the rapid development of digital image processing technology, it has been widely used in industrial and engineering fields. The target detection area, one branch of digital image operating technology, is very suitable in the bridge crack detection because of its high accuracy and constantly working.Moreover, the main work is as follows:(1)Research realized the crack identification and classification methods based on image processing technology:First, preprocessing the image samples to highlight and enhance the target part and weaken the rest, which contains image graying-. filtering and image enhancement. Second, separating the image preliminarily and weakening the interference area after image preprocessing. Then in terms of factors of connected region like area、aspect ratio and squareness author selected the suitable connected region of the extracted binary image. The next step is showing the connected regions which meeting the conditions in the original picture. After this, considering the importance of the classification of cracks, the cracks will be classified by using projection method. As to the video, processed it frame by frame by using the same method, to improve efficiency of the crack detection. Manage the video frame by frame by using the above method, and it can improve the efficiency of the crack detection.(2) Research realized the crack semi-automatic methods based on digital morphological methods:The cracks appeared on picture or a frame of video is semi-automatic measured by using digital morphological methods or manual measured to analysis of crack further, the measurement modes include measurement for images and actual measurement. Furthermore, information obtained by crack analysis and the basic information of the cracks will be stored in the database for the post-maintenance of the bridge.(3)A software of the identification and measurement of the cracks on the undersurface of the bridge that based on Visual C++is developed.