STARdbi: A pipeline and database for insect monitoring based on automated image analysis

Tamar Keasar, Michael Yair, Daphna Gottlieb, Liraz Cabra-Leykin, Chen Keasar

Research output: Contribution to journalArticlepeer-review


Insects are highly abundant and diverse, and play major roles in ecosystem functions. Monitoring of insect populations is key to their sustainable management. However, the labor and expertise needed to identify insects, and the challenges of archiving the wealth of data collected in monitoring programs, often limit these efforts. We describe a pipeline to reduce the barriers associated with curating and mining big data of insect biodiversity. The pipeline, STARdbi, includes capturing flying insects with sticky traps, scanning the traps, storing the trap-images in a public database with a web-based interface, and applying machine learning models to extract information from the images. To illustrate the insights that can be gained from STARdbi, we describe two case studies. One of them involves monitoring of circadian activity patterns of grain pests and of their natural enemies, and the other compares insect abundance, biomass and size distributions between agricultural and semi-natural habitats. We invite the community of insect ecologists to contribute to the STARdbi database, and to use its image analysis tools to address diverse ecological and evolutionary questions.

Original languageEnglish
Article number102521
JournalEcological Informatics
StatePublished - May 2024

Bibliographical note

Publisher Copyright:
© 2024


  • Classification
  • High-throughput screening
  • Insect monitoring
  • Machine learning
  • Object detection
  • Sticky trap

ASJC Scopus subject areas

  • Ecology, Evolution, Behavior and Systematics
  • Ecology
  • Modeling and Simulation
  • Ecological Modeling
  • Computer Science Applications
  • Computational Theory and Mathematics
  • Applied Mathematics


Dive into the research topics of 'STARdbi: A pipeline and database for insect monitoring based on automated image analysis'. Together they form a unique fingerprint.

Cite this