Segmentation of microscopic images of bacteria in Bulgarian yoghurt by template matching

The diagnosis of deviations in quality of yogurt is performed by approved methods set out in the Bulgarian national standard (BNS) and its adjacent regulations. The basic method of evaluation of the microbiological quality of the product is the microscopic. The method is subjective and requires significant processing time of the samples. The precision of diagnosis is not high and depends on the qualifications of the expert. The systems for pattern recognition in the most natural way interpret this specific expert activity. The aim of this report is to assess the possibility of application of a method of processing and image analysis for determination of the microbiological quality of yogurt. Selected method is template matching. A comparative analysis is made of the methods for template matching. The comparative analysis of available algorithms showed that the known ones have certain disadvantages associated with their rapid-action, the use of simplified procedures, they are sensitive to rotation of the object in the template. It is developed algorithm that complement these known and overcome some of their disadvantages.