A system developed by researchers at Rajabhat University in Nakhon Pathom, Thailand, promises to revolutionize the way tropical shrimp (Litopenaeus vannamei) are farmed in ponds.
As explained by the researchers, this innovative system uses advanced image processing and machine learning techniques to estimate shrimp weight in real-time, avoiding direct contact and reducing the risks of disease transmission. With this technological application, farmers can monitor shrimp growth from a web application, improving efficiency and profitability, especially in shrimp farming, where margins are often very tight.
The system consists of two main components: an automatic feeding tray and a weight estimator. The tray captures images of the shrimp at regular intervals, while the weight estimator analyzes these images to extract features such as area, width, length, perimeter, and posture of the crustaceans. These features are used to accurately calculate the weight of the shrimp, achieving an accuracy of 94.5%, a remarkably high figure in the field of aquaculture.
The system can operate in real conditions in ponds with turbidity and low visibility. Unlike traditional methods that require high-resolution cameras and constant maintenance, this system uses low-resolution cameras and does not need submerged electronic equipment, significantly reducing maintenance costs and facilitating its implementation in shrimp farms worldwide.
In terms of sustainability, by allowing more precise management of feeding, farmers can avoid overfeeding and food waste, which not only reduces operating costs but also minimizes the environmental impact of aquaculture by decreasing the amount of nutrients entering the pond water.
Related
