Urban Street Landscape Analysis Based on Street View Image Recognition

As part of the display of the city’s appearance, street landscape has played an increasingly important role, it is necessary to carry out systematic research on street landscape. This paper uses the Baidu map open platform to obtain the urban streetscape pictures of Dujiangyan main city, and uses ArcGIS, Python and other software to identify the panoramic pictures of the main streets of Dujiangyan downtown. From this, the visual state of street architecture, greening, sky, and water landscape elements is statistically analyzed. Through the height control of street buildings, the greening of tourist routes, the display of city skylines, and the design of water resources, Dujiangyan pays attention to the sense of change, ecology and agility in the construction of streetscapes and urban landscapes in the central city, highlighting the cultural characteristics of western Sichuan. In addition, the street has a good openness and there are differences in the landscape design among the streets. Therefore, urban streetscape image recognition can efficiently and accurately analyze the real landscape of the street, and then extract the core points of the streetscape design of the central city, which is conducive to the creation of a more regional streetscape.