Automatic Wood Defects Recognition Comparative Research

The values of a board have a direct relationship with the grading determined by the number and distribution of defects. Currently, wood defects recognition research are one of the key interests in the wood industry. In this paper, DWT and NMF are employed and we apply the decomposition to wood feature selection. The feature images by DWT can describe the characteristics and differences of the defects. Simultaneously, the wood images are decomposed by NMF. Next, we adopt DTCWT to extract the features with less redundant information from the eigen-spaces and propose a new wood defect detection method, which can effectively overcome the worse fault-tolerance and capacity of feature description. The method has been tested with color wood images. Based on visual valuation, the errors are relatively lower. After many comparative experiments, the results show the system is effectual and practical, which is not only significantly reducing the number of the errors, but also having better robust to dark spots and the analogous interferences. Besides, it can improve the performances of detection system effectively and is better than the other existing methods, having good research values and potential applications.