OpenWeedLocator OWL

OpenWeedLocator OWL detects fallow weeds using image analysis to identify and map weed populations for site-specific herbicide applications and fallow-phase weed management.

Key Features:

• Image-based Detection: Utilizes image analysis to identify low-density fallow weed populations from field images.
• Color-based Algorithms: Implements four color-based algorithms for pixel- and region-level weed discrimination.
• Algorithmic Validation: Algorithms were validated across seven fallow fields in New South Wales, Australia.
• Performance Metrics: Demonstrated average precision of 79% and recall of 52%, with individual transects reaching up to 92% precision and 74% recall.

Scientific Applications:

• Fallow Weed Monitoring: Enables identification and spatial mapping of weed populations in fallow fields.
• Site-specific Herbicide Application: Provides location data to support targeted herbicide treatment.
• Fallow-phase Efficiency Research: Supports studies aimed at reducing weed competition in moisture-limited environments to improve subsequent crop yield potential.
• Sustainable Weed Management: Contributes data for strategies to reduce herbicide use and inform resource management.

Methodology:

Images are captured from fallow fields and processed using four validated color-based algorithms.