A mango biological fingerprint anti-counterfeiting method based on Fuzzy C-means clustering

Abstract

The anti-counterfeiting of agricultural products plays an important role in protecting the rights and interests of consumers and maintaining the healthy development of the food market. Traditional anti-counterfeiting technology mainly relies on anti-counterfeiting features of packaging or labeling, which has the risk of being copied and reused. Biological fingerprint anti-counterfeiting is a method of anti-counterfeiting that takes the biological fingerprint of agricultural products as the anti-counterfeiting feature. This paper aims to take the distribution of lenticels on the surface of mango as a biological fingerprint, and propose a mango biological fingerprint anti-counterfeiting method. As the mango ripens, the peel color of mango will change significantly, which will affect the accuracy of anti-counterfeiting identification. In this paper, the images of ripe mangoes are classified by Fuzzy C-means clustering, and appropriate image enhancement technology is used to highlight the features. The results show that the mango biological fingerprint anti-counterfeiting method based on Fuzzy C-means clustering has good accuracy and robustness, and effectively reduces the impact of peel color change on anti-counterfeiting identification during mango ripening. These results support that it is feasible to use the lenticels distribution of mango as a biological fingerprint. In this paper, a computer vision anti-counterfeiting method based on lenticels distribution is proposed.

Keywords

• biological fingerprint
• anti-counterfeiting of agricultural products
• fuzzy c-means clustering
• computer vision
• mango