Skip to content

The Insect Detect system can be used for non-invasive automated monitoring of flower-visiting insects and is based on low-cost off-the-shelf hardware components combined with open-source software. Custom trained deep learning models detect and track insects landing on an artificial flower platform in real time on-device and subsequently classify the cropped detections on a local computer. The solar-powered camera trap is fully weatherproof, which enables autonomous deployment during a whole season. On-device detection and tracking can estimate insect activity after metadata post-processing.The camera trap design and open-source software are highly customizable and can be adapted to different use cases.

Left: the modular build allows for flexibility, right: view of the camera

Deutsch

On-device pipeline: The system feeds scaled-down frames (LQ) into the insect detection model. Each detected insect then gets its own unique tracking ID assigned by an object tracker and is followed as it moves. Metadata, including timestamp, insect coordinates and tracking ID, is saved together with synchronized high-resolution frames (HQ) to the Raspberry Pi’s microSD card. Optionally, the detected insects can be cropped from the HQ frames in real time.

  • Host: Raspberry Pi Zero 2 W
  • Camera: Luxonis OAK-1 (12MP)
  • Power manager: Witty Pi 4 L3V7
  • Batteries: 88Wh Li-Ion
  • Power: 5W solar panel
  • Power consumption: ~4.2 W
  • Attractant: colored platform
  • On-device processing: detection, tracking, optional: crop detections
  • Components price: ~660 €

Docs: https://maxsitt.github.io/insect-detect-docs/
Software: https://github.com/maxsitt/insect-detect
Paper: https://doi.org/10.1371/journal.pone.0295474

Contact:
Maximilian Sittinger
maximilian.sittinger@idiv.de
Developed at: Julius Kühn Institute (JKI), Helmholtz Centre for Environmental Research (UFZ), German Centre for Integrative Biodiversity Research Halle-Jena-Leipzig (iDiv)