Robot vision to automate tree pruning
Purdue University (West Lafayette, IN, USA) researchers have received $6m in grants to automate one of the most laborious and costly jobs in vineyards and apple orchards.
Each year, fruit growers spend about 20 percent of their labor costs on pruning vines and trees. Now, Peter Hirst, an associate professor of horticulture, is leading a program to develop and improve machines that use cameras and robotic arms to do the work.
Hirst received a $3m grant from the US Department of Agriculture's Specialty Crops Research Initiative, with another $3m in matching funds from industry partners and institutional funds.
The first part of the project will involve improving the prototype of an automated grapevine pruner developed by Vision Robotics (San Diego, CA, USA) a partner in the program. A tractor pulls the machine over grape vines while cameras capture images of the vines and a computer tells robotic arms where the vines should be pruned.
Presently, the pruner is at the prototype phase, but Hirst said the aim of the program was to move that machine into a commercially available device that would be cost-effective for grape producers.
The group will also start work to model trees, examine how they should be pruned, and to begin the design of a robotic pruner. Building the robotic pruner for apple trees will have to wait until the researchers receive a further grant.
To develop the system, Hirst will work with Johnny Park, a Purdue research assistant professor of electrical and computer engineering, and Avi Kak, a professor of electrical and computer engineering, who will build digital 3-D models of trees and vines and design software.
"The challenge is to accurately construct 3-D models of trees and vines and program that in such a way that the computer can determine the optimal points to prune. Our goal is to make this work at a speed that makes sense financially for growers," Park says.
Related articles on agricultural vision from Vision Systems Design.
1. Asparagus harvested by robotic vision system
European engineers are developing a continuous, fully automatic vision-based robotic system to harvest both white and violet asparagus. The system is claimed to be the first that will be able to do so.
2. Hyperspectral imaging system detects defects on apples
A researcher at the University of Maryland (College Park, MD, USA) has shown that a hyperspectral image processing system can identify accurately 95 percent of the defects on the surface of Red Delicious apples.
3. Vision system sorts tomato seedlings
A plant growing company sought help from an agrifood research organization in the Netherlands in developing an automated vision system that could classify tomato seedlings by mass.
Vision Systems Design magazine and e-newsletter subscriptions are free to qualified professionals. To subscribe, please complete the form here.
-- Dave Wilson, Senior Editor, Vision Systems Design