Design of real time image processing machine for carrot classification

Kasınhanı district of Konya province has the greatest carrot production in Turkey. By the year 2017, Konya Province has approximately 46.5% of carrot production areas and 59.7% of total production. There are several washing and packing facilities in the region. These facilities show totally similar features and fully satisfy the needs of the region. Carrots coming from the washing pools come firstly to the mechanical grading machines and then to the packing department or directly to the packing department in some facilities. Grading and packing processes are carried out manually in these facilities. The classification efficiency of mechanical classification machines is known to be insufficient. In this study, mechanical, electronic and software sections of the real-time image processing machine are explained. The system was composed of a belt conveyor, cameras and closed chamber to receive images, image processing and control computer and routing covers attached to servo motors. As a result of the experiments, carrot classification rates ranged from 80.14 to 100% in real-time image processing machine.