This is one of major outputs of the project – Mr Stevan Cakic just defended his Master thesis titled “Computer Vision and Machine Learning with Application to Food Supply Chain”. The work focuses on food tracking and tracing application of computer vision demonstrated for counterfeit prevention use case in wine industry.
ABSTRACT – Nowadays, we are all aware of the accelerated development of information technology and digitalization around the world. Montenegro is adapting and advancing in this segment as well. The purpose of this research is to apply computer science to the domain of digital transformation of food supply chain. This research focuses on application of computer vision and machine learning to wine supply chain from production to end consumers. The thesis addresses an implementation of the solution based on mobile and web applications whose primary purpose is product tracking and tracing in order to prevent counterfeiting in supply chain. The system detects individual product serial numbers obtained from photographed bottle labels, and this information is then used in heuristics algorithm to conclude whether it is a forgery or not. In order to precisely detect the serial numbers, a variety of image processing algorithms have been utilized including removal of shadows, horizontal lines, image normalization and the like. Finally, conclusions, potential challenges and possible future research are discussed at the end.
