Identification and location of corn plants with machine vision

Jiancheng Jia, Purdue University


Maize is one of the most important crops in agriculture. The identification and location of maize plants is an important issue for the development of potential automatic machines. One such machine would apply pesticides or special biotech chemicals quantitatively and precisely from a moving vehicle to promote production in fields and/or greenhouse operations. Machine vision is a potential technology that could be used to identify and locate a specific part of the maize plant in a complex bio-environment. A feasibility investigation of using machine vision technology to locate maize plants was conducted. Two approaches were proposed to locate the center and height of the maize plant using a camera from top view and side view of plants, respectively. Maize plant images under natural bio-environment were analyzed in the visual band of the spectrum using image processing techniques. An algorithm was developed to locate the center of the maize plant from a top view. The center of a plant was located by sequentially executing two image processing algorithms. Maize plants were first identified using a main vein detection algorithm by detecting a local feature of maize leaves, leaf main veins, based on the gray level difference between the main veins and leaves. Then the center of the plant was located using a center locating algorithm by tracing and extending each detected vein line and evaluating the center of the plant by calculating the average of intersection points of those lines. The performance of the algorithm was evaluated using both greenhouse and natural field images. The plant height measurement and associated problems with incorporating machine vision system on a moving vehicle are also briefly discussed. The results of this research have shown that the center of maize plants can be identified and located with image processing techniques by detecting a specific leaf feature. The experimental results show the usefulness of the algorithm for the machine vision applications related to maize plant identification and location in a bio-environment.




Krutz, Purdue University.

Subject Area

Agricultural engineering|Artificial intelligence

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server