|
Image Recognition (IR) based pest counting
We developed an AI-based decision support system for polyhouse capsicum cultivation that integrates microclimate monitoring, short-term forecasting, and image recognition-based pest counting. The system continuously records temperature and humidity inside the polyhouse and predicts the next 2–3 days climate trend using historical data. Farmers or field staff can capture plant images through a mobile application, which automatically detects and counts pests on capsicum plants using computer vision.
By combining the current pest population with predicted temperature and humidity, the system estimates whether pest infestation is likely to increase or decrease in the coming days. The output is provided as a simple risk advisory (Low / Medium / High), enabling timely pest management decisions and reducing delayed intervention. This integrated approach supports early warning, improves spray planning, and promotes data-driven IPM practices in protected cultivation.
|