Real time feature extraction and Standard Cutting Models fitting in grape leaves

In food industry, a continuous effort is made to automate the production lines and, in many cases, it is desired to automate the plant leaves processing. In this paper we present a real-time vision-based system for main features extraction of grape leaves and for positioning and fitting of Standard Cutting Models (SCMs) onto leaves’ profiles. The system comprises of a real-time software application connected to a fixed camera that captures still images of grape leaves, as they are moving on a conveyor belt. Our research focused on combining existing and novel, application specific, algorithms for image processing and developing the software application in visual C++. The aim was to create a reliable system that would process grape leaves automatically and in real time, reducing the wastage and enabling the optimization of the whole process in an industrial production line. Experimental results are also presented, in order to verify and confirm the efficiency and effectiveness of the proposed system.

Research highlights▶ Vision-based system for automation of stuffed grape leaves production. ▶ Feature extraction of grape leaves in real-time (under 1 s processing cycle) using a camera above a moving conveyor. ▶ Fitting of Standard Cutting Models (SCMs) onto grape leaves’ surface minimize wastage. ▶ Experimental results verify the quality and reliability of the “NtolcutPT” software application.