Development Of Wood Defect Detection System Based On Image Analysis

The traditional wood detection method is usually manual detection.Due to the problems of low efficiency and unstable detection quality,it is urgent to explore a new non-destructive testing method for wood defect detection.In this paper,by using computer image processing and analysis technology,proposed a set of automatic detection system for wood surface defects.On the basis of consulting a large number of references and learning relevant theories,is presented a wood defect detection system on image analysis technology.The system mainly includes four modules:wood image preprocessing module,wood defect detection module,database module and mailbox prompt module.The system first preprocesses the input wood image,including:grayscale processing,bilateral smoothing filtering and histogram statistics.If it is preliminarily determined that there is wood defect in the image,then adopt defect detection.The preprocessing wood image is segmented by threshold,edge detection and contour extraction to obtain defect location information.Then the information is stored by field matching through database.Finally,the system uses the simple command line(Smtp)to use the default command line,combined with specific characters in the database,to send the wood defect location information,the original image and the test result map to the target mailbox.The experimental results show that the system developed by image analysis technology can quickly and effectively extract the surface defects of wood surface.Since the system automatically selects the threshold by using the Otsu algorithm that introduces the neighborhood gray-scale average information,the binary defect image segmentation and extraction is performed on the wood defect region.This improved Otsu algorithm improves the accuracy of defect region segmentation.