NIR Spectroscopy Based Milk Classification and Purity Prediction

Milk adulteration is increasing day by day and people are finding newer ways and chemicals to adulterate milk. Water is the most common milk adulterant. Adulterated milk can impair the functioning of various organs of the body, causing heart problems, cancer, and in extreme cases, even death. Monitoring of milk quality is necessary for sustaining human health. To overcome this problem, a non-destructive system is proposed that detects milk purity using NIR Spectroscopy and machine learning. Dataset is created from cows and buffalos milk samples that are adulterated with varying water quantities. This dataset is used to determine if the sample is buffalo's milk or cow's milk using KNN. The acquired data is used to train model which predicts the sample purity using linear regression. Classification is done according to level of adulteration. This paper proposes a non-destructive technique to predict purity of unknown milk sample.