A Novel Approach to Food Safety Hyperspectral Imaging for Automated Adulteration Detection

Objective of this study was to design and develop new methods for detection of adulteration using Hyperspectral Data. To overcome the gaps between production, consumption, and inability of governing authorities to check adulteration are a big temptation for fraudsters to make easy money through food adulteration, at the cost of nutrition and health of consumers. In present scenario, food adulteration analysis is performed based on chemical extraction; to fill the gap our proposed system makes automatic food type and quality adulteration based on hyperspectral data in the visible and NIR spectral band. The system is cost-effective, time consuming and user friendly for the end users. The final accuracy was analysed using Support Vector Machine (SVM), K-Nearest Neighbour (KNN) techniques.