Method of Accessing the Urban Public Space from GF-2 Image by Indicator SDG 11.7.1

With the rapid urbanization, many cities have faced the problem about the reducing areas of public spaces. Many of the public spaces which are available to urban residents is compressed due to unreasonable urban planning. Among the 17 sustainable development goals proposed by the United Nations in 2015, the goal 11 is sustainable cities and human settlements, in which, the indicator 11.7.1 is the urban public space that residents can accessible. This paper used GF-2 satellite imagery to extract the information of urban features such as buildings, roads, and green spaces in Xi'an, China by means of deep learning technology. Combined with the urban population statistics, the per capita public space, green space rate and other information in Xi'an were calculated based on the SDG11.7.1 index calculation model, Compared with the 11.7.1 indicators results from the UN report, it shows that the per capita activity public space in Xi'an is at a medium level, while there is a problem of uneven distribution of urban public space. Based on the research results of GF-2 images, this paper can provide decision support for the urban planning of public space in Xi'an.