Improve the efficiency of the seedling transplanter based on machine vision

Abstract

Seedling Transplanter is an important and representative automatic agriculture machine used in greenhouse planting. During the transplanting process, it's crucial to find the disease seedling and pick it from the plug tray in order to prevent diseases and pests diffusion. In this paper, one intelligent recognition and judge system for disease seedling is presented based on Machine Vision technique. Acquire the disease seedling sample is the premise for the veracity of the vision system and this is done with the help of agriculture exports. In one transplanting step, four seedlings are extracted from the tray and taken photo in one picture. Through Image Processing, the characteristic of each seedling is obtained and used to be compared with the disease seedling samples. With the intelligent recognition arithmetic, Support Vector Machines, the vision system has more automatic and higher efficiency. © 2012 Springer-Verlag Berlin Heidelberg.