Blue cheese cut surface evaluation by images analysis

Abstract

Cheese is a food consumed all over the world. Many different types of cheese are known with variety in structure, color, characteristic flavor etc. The subject of this research is the blue cheese. It has specific smell and flavor that is related to activity of mold Penicillium roqueforti in technology process. The mold grows gradually in the process of ripening the cheese and forms a net. A good appearance and taste of blue cheese strongly depends on the even distribution of blue mold. This paper presents an image processing algorithm for evaluation the distribution of blue mold on cut surface of the cheese. The algorithm separates image into nine region of interests (ROIs) and calculates percent for areas with mold vs total area of regions. It is calculated the number of areas with mold for whole cut surface and for all ROIs. The algorithm produces information about statistical distribution of areas with mold by their size. A statistical analyzes are performed in order to evaluate distribution of mold for examined five trademarks of blue cheese. The results show that proposed algorithm determines effective distribution of mold on cut surface of the blue cheese and could be used to develop a method for automatic assessment of the blue cheese cut surface.