Abstract: In order to realize automatic identification of female and male of fragrant pear，characteristics of calyx and shape of fragrant pear were used for classification.Image of fragrant pear was collected by camera，and image was preprocessed with gray，segmentation，rotation，edge processing，contour extraction and so on.Characteristics of roundness at calyx and short and long axis of fragrant pear were obtained.Classifier model was obtained by neural network training，which was used to judge female and male of fragrant pear.The method could effectively extract characteristics of fragrant pear，and quickly identify male and female of fragrant pear，accuracy rate could be 92%.