Abstract: To study the morphology of cut stems, the captured images of cut stems were processed with an image analysis software, subjected the morphological indexes (K) of cut stems of different lengths at three specified widths to polynomial model fitting, correcting and analyzing by Matlab, and a morphological index model for cut stems was established by the division of model region combined with the morphological appearance characteristics and morphological index of cut stems. The results showed that:1) The Otsu threshold segmentation based on gray histogram enabled the extraction of cut stem image and ensured the accuracy of measurement. 2) After correction, the fitting equation presented smaller root mean square errors (RMSE=0.008), higher determination coefficient (0.995), and higher model accuracy. 3) The cut stems falling in regions Ⅰ A and ⅡA (0.039 < K < 0.169) in the model were in the shape of broken strands, in regions ⅠB, ⅡB, ⅢB (0.067 < K < 0.399) were strand-like, in regionsⅡC (0.095 < K < 0.255) andⅢC (0.255 ≤ K < 0.577) were strand-like and flake-like respectively, and in regionsⅡD and ⅢD (0.105 ≤ K < 0.772) were in the shape of flake; 4) The morphological patterns of six cut stem samples distinguished by the established model accurately reflected their morphology, and the quantitative and qualitative analysis of cut stem morphology was realized.