An Approach of Flour Adulteration Detection in Turmeric Powder Using Computer Vision

Turmeric is one of the major ingredients of Asian cuisines. It is majorly used in powdered form to add color, flavor, and taste to food. It is also an integral part of the pharmaceutical industry for its biochemical components which provide healing power to different diseases. Powdered turmeric is often adulterated. Adulterants having the same texture are added with powdered turmeric to gain more economic profit. Sometimes toxic and carcinogenic dyes are also used to enhance appearance. Our current work is focused on finding flour adulteration in turmeric powder by computer vision. An in-house image database has been used for the experimentation. The data analysis and classification tasks have been accomplished with convolutional neural network (CNN). The presented framework predicts the flour of adulteration in turmeric samples with an accuracy of more than 98%.