Under the background of the rapid development of science and technology, technologies in many fields are advancing accordingly. The same artificial intelligence and computer vision technologies are also constantly being updated. Similarly, in vision On the basis of, positioning and mapping (technology is the core technology for mobile robots to complete the intelligentization) has also had a great response in the academic and industrial circles, and has attracted the attention of most scholars. Today, the construction of smart tourism is in a prosperous stage. Based on the big data of rural tourism construction in a certain province, this article mainly uses the SLAM method that relies on visual characteristics under dynamic scenes to study the construction of rural tourism platform. In this paper, considering the influence of many factors, based on the ORB-SLAM2 system, a visual SLAM system is established. At the same time, the mathematical model of the SLAM system is established and used RGB-D The depth camera is used as the input sensor, the visual odometer locates the camera, the nonlinear optimization optimizes the data association, the closed-loop detection completes the camera repositioning, the map construction completes the environment description and other multi-module collaborations, which together complete the simultaneous positioning of the camera and the map Build work.