Towards a standardized framework for AI-assisted, image-based monitoring of nocturnal insects

Abstract

Automated sensors have potential to standardize and expand the monitoring of insects across the globe. As one of the most scalable and fastest developing sensor technologies, we describe a framework for automated, image-based monitoring of nocturnal insects—from sensor development and field deployment to workflows for data processing and publishing. Sensors comprise a light to attract insects, a camera for collecting images and a computer for scheduling, data storage and processing. Metadata is important to describe sampling schedules that balance the capture of relevant ecological information against power and data storage limitations. Large data volumes of images from automated systems necessitate scalable and effective data processing. We describe computer vision approaches for the detection, tracking and classification of insects, including models built from existing aggregations of labelled insect images. Data from automated camera systems necessitate approaches that account for inherent biases. We advocate models that explicitly correct for bias in species occurrence or abundance estimates resulting from the imperfect detection of species or individuals present during sampling occasions. We propose ten priorities towards a step-change in automated monitoring of nocturnal insects, a vital task in the face of rapid biodiversity loss from global threats. This article is part of the theme issue ‘Towards a toolkit for global insect biodiversity monitoring’.

Emerging technologies in citizen science and potential for insect monitoring

“…Similar advances enabling citizen scientists to use technology to fill spatial and temporal gaps could be explored for different types of insect monitoring; for example, camera traps for automatic monitoring of night-active insects are being developed. The advances are rapid—just a couple of years ago they could only identify few and distinct species [ 33 , 70 ], whereas now automated moth identification can be done for thousands of species [ 71 ].…”

“…Automated multisensor stations are being conceived for combined monitoring of multiple aspects of biodiversity. Current iterations include automatized visual monitoring, image analyses and bioacoustics monitoring [ 71 ], but could be extended to the detection of smellscapes using volatile organic compounds or malaise and pollen traps for metabarcoding [ 70 ].…”

Towards a toolkit for global insect biodiversity monitoring

“…Since insect camera designs are still in their infancy, we ultimately need, but are a long way from, defining and implementing standards for hardware, data generation and image analysis. As a first step towards achieving this goal, Roy et al [ 42 ] discuss metadata standards for hardware and image analysis. With the continued development of hardware and analytical algorithms, rigorous metadata recording will be fundamental for future trend analyses and, possibly, reprocessing of (historic) images [ 43 ].…”

“…[ 34 ] and Roy et al . [ 42 ] make important steps towards establishing standards for molecular monitoring and computer vision, respectively. There is also a need for harmonized storage of data and metadata.…”

Towards edge processing of images from insect camera traps

“…Nocturnal insects are difficult to monitor, however, camera-based light traps (Bjerge et al 2021b, Korsch, D., Bodesheim, P., Denzler, J. 2021) and advancement of standardized hardware and framework for image-base monitoring of nocturnal insects (Roy et al 2024) paves the way for increased temporal coverage and resolution in insect monitoring. Automated monitoring of moths has been evaluated by comparing traditional lethal methods with light-based camera traps (Möglich et al 2023).…”