Abstract
Accurate delineation of agricultural field boundaries is crucial for farm management, research, and policy development. However, publicly available boundary datasets are often limited in accuracy, very expensive, or use deep learning architectures that require extensive annotated data that are not available, limiting access. This project presents a simple, image-processing-based method for delineating field boundaries using openly available satellite images. Our method manipulates the images using a variety of image processing techniques and is able to generate an accurate boundary for a selected field. By providing a simple and accessible solution, this approach has the potential to transform boundary delineation practices, increasing the efficiency of farmers, advisors, and researchers. Our method’s simplicity and accuracy highlight a path from innovation to a real-world impact.
Keywords
image processing, GIS, computer vision, field delineation
DOI
10.5703/1288284318185
A simple approach to delineating field boundaries using satellite imagery
Accurate delineation of agricultural field boundaries is crucial for farm management, research, and policy development. However, publicly available boundary datasets are often limited in accuracy, very expensive, or use deep learning architectures that require extensive annotated data that are not available, limiting access. This project presents a simple, image-processing-based method for delineating field boundaries using openly available satellite images. Our method manipulates the images using a variety of image processing techniques and is able to generate an accurate boundary for a selected field. By providing a simple and accessible solution, this approach has the potential to transform boundary delineation practices, increasing the efficiency of farmers, advisors, and researchers. Our method’s simplicity and accuracy highlight a path from innovation to a real-world impact.