Multi-Lane Automated Catfish Seed Counting System Using Real-Time Operating System on Embedded Platforms

Catfish (Clarias batrachus) is a widely consumed food source in Indonesia, making its cultivation an essential aspect of the aquaculture industry. During the cultivation phase, catfish seeds, typically ranging from 2 to 10 cm in length, are sorted by size and counted. Conventional methods employed by seed sellers, including manual hand-counting, measuring cups, and weighing scales, are often labor-intensive and prone to inaccuracies. To enhance efficiency and reduce counting errors, this study proposes an automated catfish seed counting system. The system is designed to rapidly and accurately classify catfish seeds into three size-based categories while simultaneously counting them. A Real-Time Operating System (RTOS) is implemented to ensure concurrent processing, enabling the detection of seeds passing through three independent sensor lines. The proposed system utilizes three infrared sensors for seed detection, an LCD for real-time result display, and an RTOS deployed on an Arduino Uno platform. RTOS executes four tasks: three sensor reading tasks corresponding to each detection line and an idle task for displaying results. Experimental validation, conducted using 40 catfish seeds across 15 trials, demonstrates the system’s effectiveness, achieving an average counting accuracy of 97%. These findings indicate the potential of the proposed system to improve the precision and efficiency of catfish seed counting in aquaculture operations.