I3S Classic and Insect Species Identification of Diptera and Hymenoptera (Mosquitoes and Bees).
Abstract
A number of image recognition systems have been specifically formulated for the individual recognition of large animals. These programs are versatile and can easily be adapted for the identification of smaller individuals such as insects. The Interactive Individual Identification System, I3S Classic, initially produced for the identification of individual whale sharks was employed to distinguish between different species of mosquitoes and bees, utilising the distinctive vein pattern present on insect wings. I3S Classic proved to be highly effective and accurate in identifying different species and sexes of mosquitoes and bees, with 80% to100% accuracy for the majority of the species tested. The sibling species Apis mellifera and Apis mellifera carnica were both identified with100% accuracy. Bombus terrestris terrestris and Bombus terrestris audax; were also identified and separated with high degrees of accuracy (90% to 100% respectively for the fore wings and 100% for the hind wings). When both Anopheles gambiae sensu stricto and Anopheles arabiensis were present in the database, they were identified with 94% and 100% accuracy respectively, allowing for a morphological and non-molecular method of sorting between these members of the sibling complex. Flat, not folded and entire, rather than broken, wing specimens were required for accurate identification. Only one wing image of each sex was required in the database to retrieve high levels of accurate results in the majority of species tested. The study describes how I3S was used to identify different insect species and draws comparisons with the use of the CO1 algorithm. As with CO1, I3S Classic proved to be suitable software which could reliably be used to aid the accurate identification of insect species. It is emphasised that image recognition for insect species should always be used in conjunction with other identifying characters in addition to the wings, as is the norm when identifying species using traditional taxonomic keys.
02 /Specification & Classification
Click any tag to browse matching papers- Assay
- Research Focus
- Approach
Reproducibility ledger
Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.
Sign in to file a reproducibility report.
Related papers
- Insect Wing Classification of Mosquitoes and Bees Using CO1 Image Recognition.
Nayna Vyas-Patel, Sai Ravela, … John D Mumford·bioRxiv·Dec 23, 2015
- Morphological variation and strain identification of insects using wings and I3S.
Nayna Vyas-Patel, John D Mumford·bioRxiv·Apr 30, 2018
- Mosquito Wing Image Repository for Advancing Research on Geometric Morphometric- and AI-Based Identification
Kristopher Nolte, Eric Agboli, … Renke Lühken·Scientific DataPeer-reviewed·Apr 29, 2025
- Accurate identification of invasive Aedes mosquito species using low-cost imaging and geometric wing morphometrics
Felix Gregor Sauer, Hanna Jöst, … Renke Lühken·bioRxiv·Apr 10, 2026
- Inspecting Morphological Features of Mosquito Wings for Identification with Image Recognition Tools
Clinton S. Haarlem, Rutger Aldo Vos·bioRxiv·Sep 7, 2018
- Accurate detection and identification of insects from camera trap images with deep learning
K. Bjerge, J. Alison, … T. T. Hoye·bioRxiv·Oct 25, 2022