Analysis Of Impact Fracture Surface Morphology Based On Gray Level Co-occurrence Matrix

Metal materials have a wide range of applications in engineering practice,chosed as a component of the equipment, to consider a lot of factors. In the low temperature environment, ductile brittle transition temperature of the metal material being one of the factors can not be ignored. There are many methods to determine the ductile brittle transition temperature, which is the most widely used method because of its simple operation, convenient operation and little influence factors. The key point of this method is how to accurately measure the fiber section rate of impact fracture surface.Measuring methods of the fiber section rate by the traditional, the measurement accuracy being not high and the efficiency being low by human factors, can not meet the requirements of industrial production. With the development of science and technology, people begin to use computer and image processing technology to measure the ratio of fiber section. At present, the impact fracture surface is analyzed by using a single pixel gray level, and the segmentation method is used to realize the segmentation of fracture image. this method is sensitive to the gray value of individual pixels of the image, and it is easy to cause the phenomenon of false segmentation of the region. In order to solve this problem, this paper proposes a set of impact fracture surface topography measurement system based on texture analysis of impact fracture surface morphology.The system consists of three main modules: the design of impact fracture image acquisition system, feature extraction based on gray level co-occurrence matrix, the clustering of impact fracture image feature. After the acquisition of the impact fracture image, using MATLAB were simulation experiments, analysis of gray level co-occurrence matrix and key characteristics of the clustering algorithm of impact fracture region segmentation, finally in 2013 C# developed the system software part for visual studio. A clear and high quality impact fracture image can effectively improve the accuracy of image segmentation, and the image acquisition system is designed. Feature extraction and feature clustering are the core of image processing algorithm in the system. In order to make the texture feature extraction image to describe the impact fracture effectively and ensure the characteristics of impact fracture of crystalline region and other regions with high dispersion value. In MATLAB of the images selected different pixel on the distance, image gray level and window size construction of gray level co-occurrence matrix, observing and analysing on energy, contrast, correlation, entropy effects, the impact fracture image construction gray level co-occurrence matrix selected parameters have important reference significance. In order to follow the analysis of clustering algorithm, the suitable value of 3 parameters is selected to extract the texture feature and gray feature of the image, and the feature vector space is obtained. Finally, the clustering method is used to classify the image of each pixel, and the impact fracture region is divided.The image acquisition system, based on the gray level co-occurrence matrix to extract texture features and clustering method are introduced in the project. The function of software system is introduced, measuring the fiber section rate of the impact fracture surface. Finally, it points out the direction of the improvement of the impact fracture analysis.