2001 Volume 63 Issue 2 Pages 92-99
This paper describes the study of a supplemental seeding method using machine vision for inspecting the seed miss in a moving tray and instantly plant the seeds into the empty cavities. The method consisted of the supplemental seeding software and hardware components. The software component had the program for detecting seeds and controlling the supplemental seeding device using image processing. The hardware composed of a supplemental seeding device, which was fitted on a commercial seeder, control box, CCD camera, a personal computer, frame grabber and an I/O board. The detecting software of the supplemental seeding method could detect the seeds in a moving tray at an accuracy of 100% for cucumber, melon, coated lettuce, eggplant and green pepper seeds, and 99.93% for tomato seeds. The image processing signals precisely controlled the supplemental seeding device such that it could plant the seeds into the empty cavities at an accuracy of more than 99% with seeding time of 0.6s per cavity. Therefore, the usefulness of the studied method was recognized for automating the production of plug seedlings.