Related Experiment Video
Updated: Oct 27, 2025
05:16
Flying Insect Detection and Classification with Inexpensive Sensors
Published on: October 15, 2014
25.4K
Prediction of Pest Insect Appearance Using Sensors and Machine Learning
Dušan Marković1, Dejan Vujičić2, Snežana Tanasković1
1Faculty of Agronomy in Čačak, University of Kragujevac, Cara Dušana 34, 32102 Čačak, Serbia.
Sensors (Basel, Switzerland)
|July 24, 2021
Summary
This study introduces a machine learning model to predict pest insect appearance using temperature and humidity data. The enhanced model improves prediction accuracy, aiding farmers in timely pest management and resource conservation.
Area of Science:
- Agricultural Science
- Machine Learning
- Entomology
Background:
- Pest insects cause significant crop yield loss.
- Traditional pest monitoring is labor-intensive.
- Early detection of pests like Helicoverpa armigera is crucial.
Purpose of the Study:
- Develop a machine learning model for daily pest insect prediction.
- Incorporate weather parameters like temperature and humidity.
- Improve prediction accuracy and reduce false detections.
Main Methods:
- Utilized sensor devices with cameras for image analysis of insect traps.
- Applied various machine learning classification algorithms.
- Extended the model to consider 3- and 5-day periods for prediction.
Main Results:
- Initial model achieved up to 76.5% accuracy in predicting insect occurrence.
- The extended 5-day period model reached 86.3% accuracy.
- The extended model significantly reduced the percentage of false detections.
Conclusions:
- Machine learning models can accurately predict pest insect appearance.
- Integrating weather data and historical periods enhances prediction.
- The proposed model offers a valuable tool for farmers to optimize pest management strategies.
