Vision analysis estimates crop yields earlier

Compiled by Photonics Spectra staff

To predict how plants will react to the harsher environmental conditions expected under climate change, computer scientists at the University of Adelaide have developed an image-based analysis method for use by plant physiologists to determine crop yield.

They have joined plant physiologists and industry partner LemnaTec to develop technology that will accurately estimate plant yield of potential new cereal varieties well before grain production. In this method, a 3-D model of the plant’s changing shape is generated based on images taken as it grows. This model would then allow physiologists to quickly measure plant structural properties.

Plants growing in the Plant Accelerator at the Waite Campus. Courtesy of the University of Adelaide.

“We want to be able to predict yield based on a collection of measurable plant attributes early in the plant’s life span, rather than having to wait for the plant to mature and then measuring the yield,” said professor Anton van den Hengel, director of the Australian Centre for Visual Technologies.

The image-based approach enables detailed, accurate and rapid estimation of large numbers of plants’ potential yields under various growing conditions, such as high salinity or drought. The technology could transform crop breeding and the agricultural industry.

The new image-based analysis will be incorporated into the Plant Accelerator at the university’s Waite Campus. Opened last year, it houses more than 1 km of conveyor systems that deliver plants automatically to the imaging and other stations.

The project, “Improving yield through image-based structural analysis of cereals,” is funded by the Australian Research Council Linkage Projects.

LemnaTec will help commercialize the technology.