Image Analysis Algorithm Research And Tomography System Development Of Wood CT

Wood components were prone to decay,insect eyes,cracking and other defects in the outdoor environment for a long time,which seriously affects the load-bearing life of wood structure buildings.Traditional inspection methods,such as growth cones and impedance meters,had the disadvantages of damaging the internal structure of wood construction materials and low inspection accuracy.Computed tomography(Computed Tomography,CT)technology could clearly show the tomographic image of the object to be tested,and had been widely used in medicine,industry and other fields.At present,medical CT detection object is human,industrial CT detection chip and mechanical parts,and the use of image analysis softwwere is mostly equipment comes with or developed by third-party companies for a certain function,detection object.The results of the survey show that there was no professional equipment for CT scanning of wood structures at home and abroad,so it is important to study the CT equipment and image analysis softwwere suitable for detecting wood structure building materials.In order to better solve the problem of nondestructive testing of wood construction materials,this thesis was based on the wood construction tomography platform developed by the participating team,and the projection data were collected with wood construction materials,cedar wood,as the experimental object.Combined with the actual needs of inspection,a CT tomography imaging system for wood was developed using MATLAB GUI.The main research works in this thesis were:(1)Construction of the inverse projection image reconstruction algorithm,ButterworthSinc filter selection.The algorithm reconstructs the image intuitively,clearly and with better effect.(2)Using MATLAB combined with custom and existing wrapper functions to develop wood 3D CT image reconstruction algorithm to complete surface-drawing and body-drawing3 D image reconstruction,as all-round observation and study of wood internal structural features,which solves the problem that traditional manual inspection can only detect wood surface.(3)Develop a wood CT tomography imaging system,which includes five modules,including wood CT image pre-processing module,wood CT image three-view module,wood CT image segmentation module,wood CT image three-dimensional reconstruction module and wood CT image measurement module.Seven interfaces such as human-computer interaction start-up interface,three-view interface,three-dimensional image interface,and two-dimensional image measurement interface were developed.(4)In the developed system,to solve the problem of under-segmentation and oversegmentation of the reconstructed images,this thesis first equalizes the images with histogram,and then uses the maximum inter-class variance method,iterative method,and maximum entropy value method for experimental comparison,and concludes that the maximum entropy value method had the best effect.And in the developed system,the maximum entropy value method was used to segment the reconstructed equalized image.For the problem of unclear edges and blurred contours of segmented images,the existing Prewitt,Roberts,Log,Sobel and Canny edge detection methods were experimentally compwered in this thesis,and the results show that the Canny method had the best effect,and the final system uses the Canny method for edge contour extraction of segmented images to achieve the best results.(5)In this thesis,the dimensions,perimeter,area and pinch angle of the region of interest(Region of Interest,ROI)were measured,and the actual size of each pixel representative was derived based on the relationship between the field-of-view diameter and the reconstructed image size.The measurement results show that the absolute error between the computer measurements and the theoretical values measured manually using vernier calipers was <±1mm,and the relative error was <±0.060,which had a high accuracy.The system test results show that the system had excellent performance and meets the existing testing requirements,and could provide strong technical support for the classification and grading of wood construction timber and the protection and repair of wood structure ancient buildings.