Application Of Digital Image Processing In Structure Detection

Measured dynamic displacement data of a structure is valuable in dynamic monitoring, detecting and analyzing. But at present, real-time monitoring and detecting of dynamic displacement of large engineering structures is a difficult problem, due to the limitation of existing measurement device and technique. As for monitoring and detecting dynamic displacement, because of giant information, complexity of measure apparatus and data processing, expensive apparatus and strict working conditions of high and new technology, there are many problems in achieving automatic processing and real-time output, which is in a stage of development. It is necessary to develop a real-time dynamic monitoring system which can be directly applied to engineering practice to achieve good results. For this system, a distant range, high-accuracy and non-contact measuring method is adopted, and automatic tracking, real-time processing data and output can be achieved. In this paper, the implementation of the real-time dynamic displacement monitoring system based on image processing technique is discussed.According to principle of image formation of the ideal optical imaging system, integrating of existing basic principles and methods of the camera calibration, and considering axisymmetric and non-axisymmetric distortion, we put forward a automatic calibration method based on region growing and surface fitting. Cubic polynomial is adopted in the surface fitting. It can be solved simply and has high precision. And it can overcome the shortcomings of the traditional calibration method whose calibration process is inefficiency and takes time. Then the corresponding relationship between the object space and pixel coordinate system is determined. So the actual displacement in the object space coordinate system is obtained by measuring the target image pixel displacement.The scenes are more complicated in the fields. To achieve the goal of real-time dynamic tracking, the algorithm should be simple and efficient. In order to ensure the stability and precision of tracking, it is not appropriate for target recognition using the traditional centroid and target identification tracking algorithms. So a window-centroid tracking algorithm is put forward. Its identification accuracy is high, and robustness is well. Then the detection of dynamic displacement can be achieved by the centroid searching method. That is, first an artificial target is fixed on the measured structure, and then through the pixel displacement increment at the marker centroid corresponding to deformation of the image and deformed image, the displacement of the target can be measured. From the test and field test data show that this method can be achieved sub-pixel accuracy level.In the Windows XP, Visual C++ 6.0 programming language is used to develop the dynamic displacement monitoring system based on image processing technique, which can achieve the structure's static and dynamic measurement. Through testing and engineering applications, its reliability and effectiveness are proved. It can be achieved by a variety of distance and measurement range monitoring.