作者: F. Mendoza , J.M. Aguilera
DOI: 10.1111/J.1365-2621.2004.TB09932.X
关键词:
摘要: ABSTRACT: A computer vision system was implemented to identify the ripening stages of bananas based on color, development brown spots, and image texture information. Nine simple features appearance (L*, a*, b* values; area percentage; number spots per cm2; homogeneity, contrast, correlation, entropy texture) extracted from images were used for classification purposes. Results show that in spite variations data color appearance, a technique is as good professional visual perception. Using L*, bands, percentage, it possible classify 49 banana samples their 7 with an accuracy 98%. Computer shows promise online prediction bananas.