The sky view factor (SVF) has been recognized as an indicator to evaluate the openness of streets in the field of urban planning. It represents the ratio of the visible sky area to the total sky area at one point in space. However, due to the time-consuming and laborious acquisition of data and manual detection in traditional measurement methods, the SVF measurement in large-scale space has been greatly restricted. With the development of street view images (SVIs), some SVI services provide panorama data of the urban street level that can be used to estimate the SVF. In this research, we developed a method to measure street-level SVF based on semantic segmentation processing to extract sky area data from SVIs and estimated the fisheye photographic-based sky view factor (SVFf). Comparison with the previous research proves the reliability and efficiency of the SVF value estimated by this method. We further generated street-level SVFf maps, which served as a design base for creating more comfortable pedestrian street spaces. In the future, using our method, we can evaluate the urban thermal environment more comprehensively and accurately, and propose more targeted urban planning measures to alleviate the urban heat island effect.