“As a new type of geographic big data for perceiving the material environment of cities, street view images have made new progress in the application research of urban renewal. Experts have proposed the Cross-C2PO model, which can unify change detection and temporal decomposition tasks, effectively introducing existing image semantic segmentation models to achieve street view semantic change detection. On this basis, a perception analysis method for the dynamic degree index of urban renewal was constructed, with the monitoring of renewal changes in the main urban area of Guangzhou from 2013 to 2019 as the goal. A comprehensive perception of changes in the panoramic view of the street scene was carried out, and the dynamic degree mapping of urban renewal from four perspectives of front, back, left, right and left was realized. The distribution of urban renewal and the intensity of physical environment changes were intuitively displayed, providing innovative methods and case studies for the combination of street view and computer vision intelligence applications.”