White bread porosity evaluation by image processing with HisMedian algorithm
Journal "Fundamental Sciences and Applications"


bread porosity evaluation
image processing


This paper presents an application of algorithm HisMedian in the field of white bread porosity evaluation. There are proposed two modifications of algorithm HisMedian. The algorithm HisMedian and its modifications are implemented on Java. The images of examined bread are processed with proposed thresholding algorithms and a ratio of white pixels to all pixels is used for correlation analysis versus bread porosity measured by standard physicochemical procedure. The highest correlation coefficient is 0,93. It is proposed an equation for bread porosity evaluation using ratio of white pixels to all pixels for binary images which are produced by algorithm HisMedian and its modifications. The lowest absolute error for bread porosity calculated with proposed equation, is 1,83. The results show that the algorithm HisMedian and its modifications could be used for computer evaluation of white bread porosity.