Research On On-Line Detection Technique Of Wood Surface Defects Based On Machine Vision

The detection technique of wood surface defects based on machine vision technology and pattern recognition theory have many advantages, such as non-breakage, rapidity, accuracy, economically, and so on, It plays an important role to rank the lumber classes automatically, to improve the commodity value of sawtimber and accelerate the automation of wood processing.In this paper, Taking wormhole, dead knot and live knot three kinds of typical wood defects as research objects, an intensive study of the machine vision methods to wood surface defects are made. The main contents of the research include wood surface image preprocessing, wood segmentation, feature extraction and the identification of defects.Image preprocessing is the first step which enhances image and eliminates noise according the features of wood surface image. Image segmentation is the key stage in the detection of defects in images of wood surface, in view of the deficiency of traditional methods such as Ostu algorithms and Renyi entropy algorithms, As wood defect is natural texture stuff, wavelet reconstruction method is presented, The texture image is decomposed by using wavelet base function in terms of the optimum decomposition levels, and the restoration image can be reconstructed by selecting the smooth subimage or detail subimages at best resolution levels. The homogeneous texture pattern can be effectively removed and only local defects are preserved in the restored image. And simultaneously using morphology as the tool which has strong operation function processes the segmented images. After post-processing, it strengthened the invisibility and integrity of the segmented images and enhanced the precision of defect extraction.To identify the wood defects, the defects are described from two aspects, the texture features (11 grey matrix parameters) and geometrical characteristics (elongation and degree of rectangle).Using BP neural network classifier to identify the defects, the correct rates of pattern recognition achieved 90%According to machine version technology, The experiment result proves that it is an effective way to solve the segmentation and identification of wood surface defects by texture features of wood surface defects images.