Advanced field sensing for improved cotton management

This project will develop a novel suite of automated machine vision tools that reduce reliance on workforce labour for agronomic scouting to potentially increase productivity and improve crop management decisions involving pests, crop growth, crop rotations, patch management and irrigation scheduling. The project will be delivered by engineers at UniSQ, and agronomists, entomologists and pathologists at QDAF. The first 15 months will explore tools that could be used for IPM and agronomic decision making, and preliminary evaluations of machine vision will be made for: -          automated beat sheet sampling assessment that auto-categorises key pests and beneficials for informed IPM decision making -          assessment of plant factors used in agronomic decision-making e.g. leaf area index, height, flower and boll counts using new smartphone 3D colour imagery sensing technology -          GIS spatial detection of key weeds and diseases (e.g. feathertop Rhodes grass and wilt diseases) that have prolonged lead times or require spatially targeted control actions repeated over time using new high resolution satellite and aerial imagery -          irrigation scheduling using new in-field thermal cameras that detect canopy temperature throughout the crop lifecycle. After the feasibility and impact of each tool is evaluated, ranked and a pathway for development identified, at least two of the tools will be refined and implemented in software for commercial adoption. These will be selected following consultation with industry end-users, a steering committee and potential commercial partners. There is potential for these tools to be stand alone and used to inform typical decision making or designed to inform existing management packages that rely on bio-factor inputs (e.g. CSD’s CottonTracka).