DEVELOPMENT OF INTELLECTUAL SUPPORT’S ALGORITHM AT MANAGE DECISIONS MAKING IN THE FEEDING PROCESS BASED ON A MACHINE VISION SYSTEM’S ON A FARM

Abstract

Analysis of information obtained from open sources has shown that feed’s abnormal overspending is up to 10% and occurs in 5% of cases. It also means that the farm loses the same share of milk yield, and with it profit. This problem solution can be achieved in various ways due diversity of both hardware (vision system) and software (methods and algorithms). It is possible cows’ feed on dairy farms malnutrition reducing with the help of software and hardware complex (PAC) being developed. This PAC will determine the remaining feed volume on the feed table’s distribution by the 3D vision system of obtaining data use analyzing. Based on this analysis, the PAC will draw conclusions and make recommendations to the farm workers, thereby feedback for the farm feeding system providing. To do this, it is necessary a vision system to choose, an algorithm for obtaining a point cloud to implement an algorithm for finding reference points and gluing images into one, so as an device algorithm for orientation in space to select and implement, an algorithm for calculating feed volume and hardware to develop and laboratorial and operational tests to conduct, as well as a database for development of recommendations to create. Two vision systems are considered for use - a stereo system and a ToF camera, and an algorithm finding for reference points based on 3D data is described.