Fig. 3

Comparison between the images in the training set and the images photographed from the fecal microscope slides. a, b, c: The plant cell images collected from Ding23 and Zhou24, which was set as training set of the deep learning model; d, e, f: the images photographed from the fecal microscope slides in current study, which were recognized by the deep learn model; a, d: Clematis terniflora; b, e: Populus davidiana; c, f: Polygonum sieboldii.