Countermeasures for The Development of AI in The Field of Protection and Utilization of Architectural Cultural Heritage

This paper discusses the relationship between AI and architectural cultural heritage, the potential of “AI+architectural heritage”, and analyzes the key breakthrough technologies of AI in the field of architectural cultural heritage and the safeguard measures for the development of intelligent architectural heritage from the aspects of survey and mapping, identification and evaluation, protection and design, repair and construction, operation and maintenance management, public participation, cultural dissemination, etc. according to the basic process of architectural heritage protection and utilization. This paper designs a shot boundary detection model for intangible cultural heritage video based on 3D CNN. The model is divided into two parts. The first part of the video frame feature representation utilizes the high-level output of the convolutional neural network. By calculating inter frame differences, a large number of non shot boundary frames can be eliminated; The second part uses a three-dimensional convolutional neural network to identify cuts in candidate boundary frames. The experimental results show that the detection accuracy of the cut lens based on this model is improved by more than 10% compared to traditional methods.