Flat-top beams are widely used in laser micromachining. Evaluation of the geometric morphology and uniformity of spot energy concentration area is important to judge the availability of beams. Obtaining the contour of energy concentration area of spot is a necessary step for analyzing geometric morphology. Routine methods are difficult to obtain the boundary surrounding the energy concentration area of spot as the edge features of flat-top beam are unobvious. In this paper, taking the flat-top beam in the laser homogenization system of laser lift-off equipment as an example, an adaptive contour extraction algorithm is proposed to obtain the contour of spot, which uses an adaptive gray threshold to divide the foreground and background, and then construct a convex hull of the maximum connected area to obtain the boundary. Geometric morphology indexes have consistent results between measured spot with noise and ideal spot image without noise based on this algorithm. The comparison results between the measured spot images with noise and ideal spot images show that RMS of light intensity has no significant difference between them. In this paper, based on spatial gray texture features from gray co-occurrence matrix, is proposed to evaluate the uniformity of flat-top beam, which has significant difference between the measured spot and ideal spot, and is helpful to evaluate the quality of flat-top beam.