An urban green land vegetation information datamation accurate determination method comprises the steps of collecting high-resolution panchromatic images, hyperspectral images and thermal infrared images in a target area, obtaining approximately continuous spectral reflectance of different plant species, enhancing the ability to detect ground feature attribute information, and combining texture features of the high-resolution panchromatic images to accurately position and recognize plant contours, and detecting the plant temperature through thermal infrared data. The method can determine and identify plant species types, obtain geographical position, area, chlorophyll, carotenoid, anthocyanin and temperature information of urban green land plants, and also can obtain crown projection areasand crown projection perimeters of trees and shrubs. An object-oriented greenbelt vegetation accurate measurement method is proposed for the first time, and urban greenbelt information in a large-area range can be rapidly and accurately measured. Compared with most inversion models based on pixels, the method has the advantages that the calculation result is faster, more accurate and more comprehensive.