Research on the intelligent identification method of wallpaper labeling in the interior decoration

In order to realize the intelligent identification of wallpaper labeling by using computer vision method, an analysis method of wallpaper labeling texture is proposed based on gray level-gradient co-occurrence matrix (GLCM). First, it determines the key parameters of matrix, makes up suitable GLCM, which the feature parameters is extracted; next, it completes the feature selection of the fourteen parameters in order to reduce the computation and improve the describing ability; last, it designs the probabilistic neural network classifier to accomplish classify experiments, whose overall recognition rate can reach 89.6%, which suggests the method is effective and basically feasible.