Detection of Cumin Seeds Adulteration Based on Thermal Imaging Using Deep Learning Technique

Abstract

Food adulteration refers to the addition of artificial or poor-quality substances to a food product. Since the start of mankind, food adulteration has been a problem that artificial intelligence has successfully detected. In addition to lowering food quality, it also has several adverse health consequences. Two spices that are often used in the food industry are cumin and fennel, although they can be adulterated. Using deep learning, this research established a convolutional neural network (CNN) architecture that can identify between a food product and additional adulterants. This technique accurately detects 95.5% adulteration in cumin and fennel seeds. Key Words: Food adulteration, thermal imaging, spices, image classification, CNN, inception V3 algorithm.