Social progress has prompted people to gradually improve the quality of life, and the requirements for the living environment are also higher and higher, and the housing interior design with artistic style has attracted widespread attention. In this paper, HSV color model and GLCM algorithm are used to extract color features and texture features of housing interior design style images, and then combined with local sorting difference refinement algorithm to obtain local features of housing interior design style images. After obtaining the two features of housing interior design styles, a simple Bayesian classifier of machine learning algorithm is constructed to recognize and classify housing interior design styles. In order to enhance the accuracy of style identification and classification, an adaptive majority voting decision fusion algorithm is introduced to distribute the weights of the plain Bayesian classification results and output the optimal weights to improve the accuracy of housing interior design style identification and classification. The average accuracy of color features extracted by HSV color model is up to 91.05%, and the classification accuracy of LSDRP algorithm is 98.39% when local features are extracted. Compared with the TSVM model, the OA, AA and Ka indexes of this paper’s method are improved by 20.79%, 29.03%, and 27.39%, respectively, when performing housing interior design style identification and classification. The use of machine learning algorithms can realize the accurate identification and classification of housing interior design styles, which provides a reference for improving the level of housing interior art atmosphere design.
