Wood recognition based on grey-level co-occurrence matrix

By reason of the convenient obtaining of wood stereogram images, it's suitable for us to apply them to the application of wood recognition. In order to extract features from the wood stereogram images, gray level co-occurrence matrix (GLCM) was used to statistic texture features. Under the image resolution of 100*100, four directions, i.e. 0°, 45°, 90°, and 135°, were severed as the generated pixel directions of GLCM. Besides, providing the pixel interval with 4 and gray level with 128; Also six features, Energy, Entropy, Contrast, Dissimilarity, Inverse Difference Moment, and Variance, were used as classification features in the experiment. According to the experiment of the wood recognition, about 91.7% recognition rates were acquired through feature extractions of 24 wood species, and 480 samples, and the use of the SVM classifier. The experiment results showed that it was feasible to apply the six proposed features of GLCM to the wood recognition, and they can finish the task effectively.