Eyes on nature: Embedded vision cameras for multidisciplinary biodiversity monitoring

Darras, Kevin F.A.; Balle, Marcel; Xu, Wenxiu; Yan, Yang; Zakka, Vincent G.; Toledo-Hernández, Manuel; Sheng, Dong; Lin, Wei; Zhang, Boyu; Lan, Zhenzhong; Fupeng, Li; Wanger, Thomas C.

doi:10.1101/2023.07.26.550656

Cited by 3 publications

(1 citation statement)

References 87 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A range of different sensors can be used for automated insect monitoring [9]. These include acoustic [10] and opto-electronic sensors [11–13], as well as cameras [14–20]. Several low-cost DIY camera trap systems use scheduled video or image recordings, which are analyzed in subsequent processing steps [14, 15].…”

Section: Introductionmentioning

confidence: 99%

Insect Detect: An open-source DIY camera trap for automated insect monitoring

Sittinger,

Uhler,

Pink

et al. 2023

Preprint

View full text Add to dashboard Cite

Insect monitoring is essential to design effective conservation strategies, which are indispensable to mitigate worldwide declines and biodiversity loss. For this purpose, traditional monitoring methods are widely established and can provide data with a high taxonomic resolution. However, processing of captured insect samples is often time-consuming and expensive, which limits the number of potential replicates. Automated monitoring methods can facilitate data collection at a higher spatiotemporal resolution with a comparatively lower effort and cost. Here, we present the Insect Detect DIY (do-it-yourself) camera trap for non-invasive automated monitoring of flower-visiting insects, which 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. Field deployment of the solar-powered camera trap confirmed its resistance to high temperatures and humidity, which enables autonomous deployment during a whole season. On-device detection and tracking reliably estimated insect activity/abundance after metadata post-processing. Our insect classification model achieved a high top-1 accuracy on the test dataset and generalized well on a real-world dataset with captured insect images. The camera trap design and open-source software are highly customizable and can be adapted to different use cases. With custom trained detection and classification models, as well as accessible software programming, many possible applications surpassing our proposed deployment method can be realized.

show abstract

Section: Introductionmentioning

confidence: 99%

Insect Detect: An open-source DIY camera trap for automated insect monitoring

Sittinger,

Uhler,

Pink

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

Bugs and bots: how technology is changing the game in biodiversity monitoring

MacWilliams,

Kim,

Trueman

2024

Biodiversity

View full text Add to dashboard Cite

Camera traps unable to determine whether plasticine models of caterpillars reliably measure bird predation

Schillé,

Plat,

Barbaro

et al. 2024

Preprint

View full text Add to dashboard Cite

Sampling methods that are both scientifically rigorous and ethical are cornerstones of any experimental biological research. Since its introduction 30 years ago, the method of using plasticine prey to quantify predation risk has become increasingly popular in the field of biology. However, recent studies have questioned the accuracy of the method, arguing that inaccuracy in inferring evidence of predator bite marks and the artificiality of models may bias the results. Yet, bias per se might not be a methodological issue as soon as its statistical distribution in the samples is even, quantifiable, and thus correctable in quantitative analyses. In this study, we focus on avian predation of lepidopteran larvae models, which represent one of the most widespread predator-prey interactions in various ecosystems worldwide. We compared bird predation on plasticine caterpillar models to that on dead caterpillars of similar size and color, using camera traps and observer posterior assessment to identify biases and quantify actual predation events. Although camera trap analyses revealed that birds respond similarly to plasticine models and dead caterpillars, suggesting no advantage in using dead prey for assessing avian predation, the results of posterior identification of predation marks showed contradictory results, making the interpretation of the results ambiguous. Observer inconsistencies in detecting predation clues on models highlight potential biases, with camera trap data showing these clues do not reliably indicate actual bird predation. Given the uncertainties and shortcomings of technological character revealed, the development of adequate monitoring methods and further research is essential to refine the assessment approach and better understand the biases in using plasticine models.

show abstract

Eyes on nature: Embedded vision cameras for multidisciplinary biodiversity monitoring

Cited by 3 publications

References 87 publications

Insect Detect: An open-source DIY camera trap for automated insect monitoring

Insect Detect: An open-source DIY camera trap for automated insect monitoring

Bugs and bots: how technology is changing the game in biodiversity monitoring

Camera traps unable to determine whether plasticine models of caterpillars reliably measure bird predation

Contact Info

Product

Resources

About