Wood surface quality is one of the important indexes to evaluate wood quality. With the development of the wood processing industry to the direction of large-scale production of mechanization and automation, people pay more and more attention to the processing quality, especially the surface defects, so the surface defect detection technology becomes increasingly important.
The wood defect is a kind of quality defect when the wood is suitable for a special purpose. It will reduce the strength of wood and affect the quality and appearance of processing and decoration. According to the national standard, wood defects are divided into 10 categories: knot, discoloration, decay, insect pest, crack, tree trunk shape defect, wood structure defect, scar, wood processing defect, and deformation. The defects of the plate mainly include knot, decay, crack, skin inclusion, insect pest, and processing defect.
The principle of using digital image processing technology to detect the surface defects of sheet metal is to take real-time mechanical photos of the plate surface with a CCD camera. After digital processing, the photos are sent to the host computer for image processing. Through parameter calculation, the features of the plate image are extracted to detect the surface defect information, and then the classification and grading are carried out. This article discusses the detection and treatment of the surface defects of sheet metal by digital image processing technology.
The schematic diagram is as follows:
Feature extraction directly affects the recognition rate of the wood defect detection system. Usually, the secondary peak is the defect location, but this is not absolute. If the histogram curve has only one peak, it may be a normal wood picture. According to the experimental statistics, when the secondary peak value is greater than 1 / 10 of the main wave peak value, the secondary wave peak represents the defect color. The first step of defect detection is to divide the image into a defect and non-defect.
Detection process: the sheet surface defect detection system based on digital image processing technology is composed of a light source, CCD camera, image acquisition card, industrial computer, and output equipment. The CCD camera connected with the industrial computer takes the wood surface image and converts the wood surface features into image signals. According to the pixel distribution and brightness information, it is sent to the industrial-grade computer through the image acquisition card, and then the industrial computer uses special processing software to store and process the collected images. Laplace operator and median filter are used to enhance and smooth the edge of the wood image.
Using image processing to detect wood defects has the advantages of high real-time speed, no damage to wood, and saving material resources. It solves the problems of large errors in manual identification of wood defects, and it is difficult to realize double-sided identification, which brings huge economic benefits to the wood industry.