Authentication of indigenous Brazilian specialty canephora coffees using smartphone image analysis.

The prevention of coffee fraud through the use of digital and intelligence-based technologies is an analytical challenge because depending on the adulterant, visual inspection is unreliable in roasted and ground coffee due to the similarity in color and texture of the materials used. In this work, a 3D-printed apparatus for smartphone image acquisiton is proposed. The digital images are used to authenticate the geographical origin of indigenous canephora coffees produced at Amazon region, Brazil, against canephora coffees from Espírito Santo, Brazil, and to capture the adulteration of indigenous samples. The results evidenced that the technology is favorable to identify the geographical origin and adulteration with multiple substances using smartphone technology. Pure coffees were adulterated with arabica coffee, spent coffee ground, low-quality Canephora coffee, coffee husks, açaí, corn, and soybean in increasing proportions of 10, 20, 30, 40, 50, 60, and 70 %. These adulterants were roasted and grounded similarly to Canephora coffees to mimetize a highly-sophisticated fraud. The images were converted into Red-Green-Blue (RGB) fingerprinting and used as analytical response to construct Data-Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) models. A total of 95 % of all target and non-target samples in the test set were correctely identified, aiding producers and consumers in ensuring accurate labeling and supporting traditional communities economically and culturally. Smartphone-based method demonstrated potential to innovate the coffee safety control representing a new analytical tecnology.