Agricultural and Viticultural Image Processing

Algorithms to improve efficiency in agtech using high-detail image processing systems to estimate yield, detect stress levels in plants and manage land using commercial, off-the-shelf hardware and consumer devices. These tools give large-scale visibility that improve productivity while saving time and money.

Competitive advantage

Simple to use image processing systems that give large-scale visibility with much greater detail than aerial imaging
Uses highly affordable and available image capture systems
Robust algorithms backed by ground-truth data

Impact

The development of more robust algorithms for processing images from tree-size to microscopic level will reduce labour requirements in vineyards and orchards, improve the reliability of image data management and provide greater insight into yield estimation processes.

Successful applications

There have been nine finished deployments and evaluations of the technology, with three in process.
Refined the data collection practices of Pernod Ricard in the Marlborough region of New Zealand.
Apple flower and canopy density mapped by SwarmFarm using algorithms
QR code based image processing tools simplifying data collection to identify leaf and trunk disease by Plant and Food Research New Zealand
Algorithms applied by Tasmanian Institute of Agriculture for flower counting and early-stage yield estimation
Algorthms being used by researchers in the Netherlands for stomata detection
Komatsu Japan is implementing patented earth-moving algorithms

Capabilities and facilities

Local and remote computing servers configured for large-scale image processing.
Field sites for evaluating mobile sensing systems.

Competitive advantage

Impact

Successful applications

Capabilities and facilities

Our partners