Shrimp AI Detection & Sizing

An AI-powered computer vision system for shrimp aquaculture that combines disease detection and body length measurement in a single pipeline.

Built on YOLO and RTMDet architectures to detect and classify shrimp health conditions — identifying disease symptoms such as white spot, black gill, and body deformities. Simultaneously measures shrimp body length with sub-millimeter precision for growth tracking and harvest optimization.

The system processes real-time camera feeds from shrimp ponds, providing farmers with instant health alerts and growth analytics. Optimized inference pipelines reduce latency from 45s to <15s, enabling continuous monitoring across multiple ponds.

Integrated intelligent contextual reasoning to achieve over 95% automation accuracy. Implemented a SQL Guardrail layer for data security and robust multi-tenancy support across farm operations.

Tech Stack: Python, YOLO, RTMDet, FastAPI, Docker, OpenCV.