The invention relates to an urban green land extracting method cooperating with a multisource remote-sensing image. The method comprises the following steps that S1, a high-resolution image and a multispectral image are collected, image registering is carried out, the registered images are stacked, and a stacked image is obtained; S2, the stacked image is subjected to image segmentation based on the ground object space-spectrum features of the stacked image in S1, and a final segmented object is obtained; S3, vegetation spectral indexes are constructed through spectral information in the multispectral image based on the final segmented object in S2, a threshold value is selected, urban green land information is obtained, and drawing is carried out. Compared with the prior art, the urban green land condition can be extracted fast and accurately, drawing is carried out, and the method can be widely applied to the fields of urban planning, urban environment assessment and the like and isbeneficial to urban planning and development of the environmental protection cause.