AUTOMATED PROCESSING OF DAIRY PRODUCT MICROPHOTOS USING IMAGEJ AND STATISTICA

Abstract

Summary. The article discusses the construction of algorithms for automated processing of microphotos of dairy products. Automated processing of micro photos of dairy products relevant in the study of the degree of homogenization. Microphotos of dairy products contain information about the distribution of fat globules in the mass fractions. Today, there are some of software products, offering image processing and relieving researchers from routine operations manual data processing. But it need to be adapted for performing the processing of microphotos of dairy products. In this paper we propose to use for processing the application package ImageJ for processing image files taken with digital microscope, and to calculate the statistical characteristics of the proposed use of the software package Statistica. Processing algorithm consists of successive stages of conversion to gray scale, scaling, filtering, binarization, object recognition and statistical processing of the results of recognition. The result of the implemented data processing algorithms is the distribution function of the fat globules in terms of volume or mass fraction, as well as the statistical parameters of the distribution (the mathematical expectation, variance, skewness and kurtosis coefficients). For the inspection of the algorithm and its debugging experimental studieswere carried out. Carries out the homogenization of farm milk at different pressures of homogenization. For each sample were made microphoto sand image processing carried out in accordance with the proposed algorithm. Studies have shown the effectiveness and feasibility of the proposed algorithm in the form of java script for ImageJ and then send the data to a file for the software package Statistica.