Automatic recognition of grapes' size level based on machine vision

作者：Li, Min; Wang, Qiaohua^*; Zhu, Jiang

来源：Journal of Food Agriculture and Environment, 2012, 10(3-4): 78-80.

• Grape
• machine vision
• image processing
• angle of curvature

In order to realize the automatic classification of grape size nondestructively, the method used for automatic grading of grapes size is introduced in this article. The grape images were collected by the machine vision device to image enhancement processing, and the grape contour was segmented by using edge detection twice, after grape contour information was extracted, the angle of curvature was used to calculate the angle of curvature of every point of grape contour, and the average value was taken as the image outline curvature angle value. The angle of curvature of grape fringe figure was selected as characteristic parameter for the size classification and the identification model of grape size. The accuracy rate of model being nearly 90%, shows that the method is feasible and the machine vision for classification of similar round string agricultural products to provide a new idea.

• 出版日期2012-10
• 单位华中农业大学