Research On Defect Detection Technology Of Aviation Component Based On Image Processing

The main purpose of the aviation component defect detection technique is to ensure the quality and performance of the aviation component products,to avoid the failure of aviation equipment and even safety accidents due to the defects of the components.In recent years,the research of automatic defect detection technology based on digital image processing has been paid more attention by researchers all over the world.However,the detection efficiency is reduced because of missing or misjudgment phenomenon which caused by the image edge blurring,low contrast and other factors.So,the main contents of this paper include image enhancement,image segmentation and defect detection,defect feature description and classification in the defect detection technique.The main works are described as following:(1)To deal with the problem that the anisotropic diffusion model on image denoising enhancement causes the edge blurring,ringing effect and so on,a new bidirectional diffusion model based on local features is presented.Firstly,on the basis of the four-orientation gradient and Laplacian operator,the flat and edge regions are separated through the deep research on the structural characteristics of the image edge.Then,according to the characteristics of each region,the presented model chooses the different diffusion methods,which can not only filter the noise but also sharpen the edge details of the image.The experimental results show that compared with other diffusion models,such as PM,CBDF and ABDF,the proposed model can effectively improve the contrast of the image and enhance the edge details of the defects.(2)Having a unitary fuzzy membership matrix structure and ignoring the relevance between adjacent pixels,the traditional fuzzy C-means clustering algorithm can not obtain the satisfactory image segmentation result.Even though some improved clustering algorithms improve segmentation result,the running efficiency is low.Therefore,a intuitionistic fuzzy C-means clustering algorithm based on hierarchy for image segmentation is proposed.First,according to the hidden information of the image,the image is divided into several different regions by introducing the hierarchical technique,and the cluster center is computed in the corresponding region.Then,on the basis of the correlation between adjacent pixels in the image,the objective function based on the relevance between adjacent pixels and the theory of intuitionistic fuzzy sets is constructed to acquire the intuitionistic fuzzy membershipmatrix.Finally,the image layer is modified continuously until it satisfies the clustering criterion.The experimental results show that compared with other image segmentation algorithms,such as FCM,FCMS and FLICM,the proposed algorithm can not only get better the image segmentation result,but also improve the efficiency of image segmentation effectively.(3)According to the structural characteristics and distribution of common defects of aviation component,the feature parameters which can accurately identify the different kinds of defects are selected,and then the binary tree classifier is constructed to classify the defects.The experimental results show that the binary tree classifier in this paper can basically identify various defects of aviation component.