Application of thresholding algorithms for brown bread porosity evaluation

Abstract

The bread is one of the most popular foods in Bulgaria. It’s quality is regulated by approved standards. This paper presents a computer based approach for evaluation of bread porosity which is one of physicochemical parameters of bread quality. The proposed approach includes image processing techniques. A Java program is developed to binarize images of bread and calculate ratio of white pixels to all (coefficient of diversity). This coefficient corresponds with physicochemical parameter- bread porosity. It is used an open-source plugin Auto_Threshold for image binarization. This plugin implements seventeen different algorithms to find global threshold value of a grayscale image. The results show that global thresholding is appropriate for evaluation of bread porosity. The correlation analysis shows that algorithm HisAnalysis could be used for fast and effective evaluation of bread porosity using image processing.
 Practical applicationsThe use of image processing accelerate the process of bread porosity evaluation. Presented research proves practical benefit to apply image processing for evaluation of physicochemical parameter- bread porosity. The results show that seven algorithms which are included in Auto_Threshold plugin and HisAnalysis algorithm are suitable for bread porosity evaluation. The fastest algorithm is HisAnalysis and it could be used in practice for fast evaluation (in real-time processing) of physicochemical parameter- bread porosity.