Extraction and Protection of Historical Architectural Features of Traditional Villages Based on Computer Vision

Traditional villages often preserve a large number of historical buildings, which contain unique cultural and economic values and are one of the important carriers of rural revitalization and cultural construction. In this paper, computer vision technology is used to model historical buildings, and fine mapping is carried out according to the model contour lines. The SIFT operator is used to extract mixed feature points from the environment and describe them. The color features of historical buildings are extracted using an HSV color space model, and the results are presented using a color histogram. The case study shows that the selected ancient villages in northern Guangdong have strong integration and integrity of the historical building results, with the southwest orientation predominating, and the vast majority of them are brick and wood structures. The color characteristics of traditional houses are warm tones, while the clan temples show the color landscape characteristics of contrasting cold and warm tones. It can be seen that the method in this paper is able to extract multiple features of historical buildings, which provides a decision-making basis for further historical building protection.