The Research On Veneer Peeling Image Defect Inspection Technology Based On PDE

As the rapid development of computer technology and artificial intelligence technology, image processing technology made great progress in recent years, Partial Differential Equations application are very extensive in the field of image processing applications. Various models for image processing in the industrial, military and medical have been developed currently, such as the human work and life have brought great convenience.This thesis briefly introduces the study backgrounds and research status of Partial Differential Equations at home and abroad and makes further studies on the methods with image segmentation of C-V model in system. Then, coupling with Technique of Painting Background method compared and studied both gray-image and color-image defect inspection. Veneer detection automatic of surface defect images by computer analysis and processing, thus reducing the operator's labor intensity and increase timber volume ratio and production efficiency, to provide the basis for veneer surface image defect detection.According to the characteristics of veneer surface defects, using C-V model to segment knot defect. For the iterative process of C-V model need to re-initialize, using anther energy function to control signed distance function, solving the problem of computational complexity. For the low efficiency of Euler method in the process of numerical calculation in C-V model., using AOS method to solve the time step limit of the algorithm. Experimental results show that this method improved segmentation speed.This thesis according the veneer surface defects variety and range wide, introduce Technique of Painting Background method. This method can reduce the characteristics between target and background, reaching the goal of narrowing the scope of identifying the veneer defect. In this thesis, inspect for the veneer surface defect based on extension of C-V model and AOS scheme coupling with Technique of Painting Background and use Matlab to make experimental verification. Experimental results show that this method can quickly and effectively segment the veneer surface defect, thereby improving the speed of the whole algorithm.This thesis analysis the live knot's boundary change slowly, using C-V model can not have the ideal segmentation, researching vector C-V model, let the image as an overall image to process, fully used the image information, resolving the problem of the live knot inspection. Using the extension of vector C-V model and AOS scheme coupling with Technique of Painting Background can both inspect live knot and identify veneer color image defect.Partial Differential Equations as an emerging technology will play an increasingly important role in image processing. This thesis based on C-V method of Partial Differential Equations, improved C-V model, coupled with Additional Operator Splitting and Technique of Painting Background. It improves existed C-V algorithm and has a great role in promoting image protection technology on veneer surface image defect inspection.