Research On Detection System Of Physical Simulation Experimental Profile On Geological Structures Based On Digital Image Processing

Physical simulation experiment on geological structures is an effective method for studying geologic structure, tectonic distortion processing and formation mechanism for geologists. At present, degree of deformation on physical simulation experiment on geological structures is often used in qualitative methods, such as visual, measuring using ruler, so there are some questions on human disturbance, measuring accuracy, and quantitative obtaining degree of deformation. The paper studies profile images of experimental flask, it detects profile by a quantitative method with digital image processing technology. The main work of the thesis is as follows.1. This paper introduces functional scheme and related parts of physical simulation experiment installation on geological structures, and describes in detail the design of mechanical system and electric control system. Automatic platform of image captured is established according to the characteristics of the system, and it can automatically capture images using digital camera controlled by computer. Control software is realized, it includes the setting of camera function module and image processing function module.2. The characteristics of profile images are that the foreground is black sand particles, and the background is common sand particles. There are plenty of deep color sand particles in the background. Therefore, mark line of the image is not clear, and the color is not uniform. These lead to the lack of feature extraction. This paper compares the differences of basic algorithms on the basis of image processing, and then finds the more effective algorithms of image processing. Firstly, it compares the difference between methods of image enhancement based on gray level transformation and histogram, and uses the method of linear transformation to enhance mark lines of the image. Secondly, it uses wavelet and spatial domain method for image filtering to pretreat image noise. This method has a favorable effect in removing point noises compared with common filtration methods, and it preserves detail informations of images.3. After gray image is converted into binary image, non-ideal binary image has noises. So, we use one improved algorithm based on mathematical morphology. Different structuring elements are used to opening and closing operation. More ideal image is obtained using this algorithm. Also, connect fracture part of the mark lines on geological images using curve fitting method. It extractes smoothing mark lines and curve inflection point of the images.The results showed that the characteristics of the image detection system are quantification and accurate detection, and it is an effective method for analyzing tectonic evolution processing of geology. This system provides accurate basic data for geologists who analyze evolution processes of geological structure.