To address the challenge of real-time dynamic monitoring of cut and fill areas during the construction phase, a method for real-time identification and volume calculation of cut and fill areas based on unmanned aerial vehicle (UAV) digital surface model (DSM) is proposed. The method first extracts the boundaries of potential cut and fill areas through multi-scale adaptive edge detection (based on an improved Canny operator); then integrates DSM elevation derivatives, terrain curvature, and roughness to construct a multi-feature decision model for determining whether the area is a cut or fill; subsequently applies an adaptive region growing algorithm to optimize the integrity of the boundaries; finally determines the reference plane based on the average DSM elevation of the regional edges, and realizes the volume calculation of cut and fill by calculating the integral of height differences between pixels in the region and the reference plane.