The automatic sorting using image processing improves postharvest blueberries storage quality

South American blueberry fruit production has been increasing over than 40% the late decade moved for international demand leading for United States of America. However, during storage and transportation stage, fungal decay may cause serious damage in batch which causes rejects and low prices in destinations. Moreover, traditional manual methods to remove diseased unities of blueberry after harvest are slow, unconfident and expensive. In this work, we propose a simple and non expensive method to be implemented to remove unities with fungal damage. It consists in computer vision algorithms to extract and select information from blueberries and implement the best classifier to segregate automatically unities with fungal decay, shrivelling and mechanical damage from health unities. It was possible classify correctly over than 96% images with fungal decay and more the 90% of blueberries with global damage (fungal decay, shrivelling or mechanic damage). The implementation of automatic computer vision systems to recognize defects in blueberries is complex due to blueberries morphology, obscure colour, little size, wax presence. It results are promissory since will allows increase export quality when will be implemented in production lines.