Digital agriculture services can greatly assist growers to monitor their fields and optimize their use throughout the growing season. Thus, knowing the exact location of fields and their boundaries is a prerequisite. Unlike property boundaries, which are recorded in local council or title records, field boundaries are not historically recorded. As a result, digital services currently ask their users to manually draw their field, which is time-consuming and creates disincentives. Here, we present a generalized method, hereafter referred to as DECODE (DEtect, COnsolidate, and DElinetate), that automatically extracts accurate field boundary data from satellite imagery using deep learning based on spatial, spectral, and temporal cues. We introduce a new convolutional neural network (FracTAL ResUNet) as well as two uncertainty metrics to characterize the confidence of the field detection and field delineation processes. We finally propose a new methodology to compare and summarize field-based accuracy metrics. To demonstrate the performance and scalability of our method, we extracted fields across the Australian grains zone with a pixel-based accuracy of 0.87 and a field-based accuracy of up to 0.88 depending on the metric. We also trained a model on data from South Africa instead of Australia and found it transferred well to unseen Australian landscapes. We conclude that the accuracy, scalability and transferability of DECODE shows that large-scale field boundary extraction based on deep learning has reached operational maturity. This opens the door to new agricultural services that provide routine, near-real time field-based analytics.
Extensive livestock production occupies 25% of the ice-free terrestrial surface of the Earth, and Australian beef production contributes about 10% of this total land footprint. Therefore, the management of cattle has major implications for natural vegetation, soils, biodiversity, and atmospheric greenhouse gases in Australia and globally. To meet global sustainability targets and consumer expectations, a variety of environmental Standards, Certification and Assurance (SCA) schemes are being developed and implemented to enable producers to verify claims relating to their products. Improved standardization and coordination are now needed to address the rapid proliferation of credentialing systems that use different frameworks, methods and levels of scientific verification. Using the Australian grass-fed beef industry as a case study, we identified the measures, metrics and methods that are currently used by SCA schemes for climate and biodiversity credentialing. From here, a co-design process with industry representatives was applied to develop recommendations for credentialing in extensive beef production, feedlots and meat processing. It was clear that the successful adoption of credentialing by beef producers will require flexible and user-friendly support tools that use scalable data sources such as existing producer records, agtech databases and remote sensing information. Substantive rewards and incentives will be required to support producer engagement with the SCA schemes. Overall, ’more needs to be done to ensure the transparency of schemes and to meet benchmarks for integrity such as determining uncertainty and support though peer-reviewed science.