Automatic System of Seeds Refined Grading Based on Machine Vision

Abstract

An automatic seed grading and selecting system based on machine vision dynamic inspection technology was developed. The working principle of the inspection system and a whole structure of the system were determined and a prototype was manufactured. Colour code sensors were adopted to trace the positions of seeds and automatically acquire the position information of each seed in a rapidly moving state. All the seeds were sorted into four grades according to their morphological characteristics. Separation of the graded seeds was realized by on-off control of magnet valves to pass or block a high pressure air flow. A programmable logic controller (PLC) was adopted to control the whole test-bed. The experiment results show that the eligible ratio of grading is over 92% and 15 seed kernels are processed per second when the belt velocity is about 0.2m/s and the air pressure is 0.85 MPa