Identification of acrylamide in fried potato crisps using image processing in wavelet domain

Abstract

Acrylamide is a toxic carcinogenic material that commonly occurs in heated starchy food items like potato crisps. Identification of such toxic chemical in food items using conventional chemical laboratory based methods is time consuming and expensive and may need specialized manpower for such laboratory testing. This paper proposes an efficient image processing based non-destructive testing method for identification of acrylamide in potato crisps. The image of the potato crisps is automatically segmented and features are extracted from the images in wavelet domain. These features are then analyzed for identification of the presence of acrylamide in these samples. The variation in the various features of the image is related to the presence of acrylamide. The experimental results indicate that the proposed method is efficient for identification of acrylamide from the wavelet features of the images. The proposed method achieves 90% of accuracy and it can be used for real time applications.