Multispectral imaging for identifying crop pest feeding damage

Overview

Crop pests such as slugs, aphids, and caterpillars cause substantial economic losses in vegetable production, yet identifying the specific culprit behind feeding damage remains a persistent challenge. This over-reliance on broad-spectrum chemical pesticides can harm non-target organisms and the wider environment. This solution applies multispectral imaging to differentiate between the distinct feeding signatures left by these three key pest groups, offering an in-field diagnostic capability to support integrated pest management (IPM) decision-making.

By providing rapid, image-based pest identification, growers and crop advisors can apply more targeted interventions, reducing unnecessary chemical inputs and improving both economic and environmental outcomes. The approach builds on demonstrated success from prior imaging-based crop monitoring work, extending the concept from disease detection to pest damage differentiation.

Technical specifications

• Imaging platform: Multispectral imaging using the VideometerLab at CHAP's phenotyping laboratory to capture spectral signatures of feeding damage
• Target species: Slugs, aphids, and caterpillars, representing mollusc and insect pest groups commonly affecting vegetable crops
• Methodology: Controlled petri dish experiments to monitor and document feeding damage per species, followed by image capture and spectral signature analysis
• Output: A reference database of feeding damage spectral signatures for each pest species, enabling future field-based identification
• Workflow stages: Pest collection and maintenance, controlled feeding trials, multispectral image capture, spectral analysis, and database compilation

Technology readiness level

The solution is at an early feasibility stage. Validation is structured across three work packages: pest collection and maintenance, controlled feeding experiments with multispectral image capture and analysis, and project management. The work leverages existing phenotyping infrastructure and prior proof-of-concept results from related imaging-based crop monitoring research, providing a foundation for further development toward field-deployable diagnostic tools.