Fast detection technology based on machine vision for hill-drop performance of precision seed-metering device

Small seeds such as vegetable seeds and flower seeds are sowed by hill-drop planter. Based on machine vision, a vision inspection system of hill-drop performance of precision seed-metering device was established for small seeds. The seed regions were marked by using labeling algorithm, and seed feature was extracted by detecting the location coordinates of the seeds. Hill spacing was calculated by measuring seed spacing first and then separating hole, and point grains was counted. The experiment showed that the results measured by the system were in agreement with manual detection. The hill spacing error within 3 mm, and point grains error less than 25%.