Fish monitoring in river by deploying IoT-UWSNs

Rivers are living libraries, their currents holding the stories of countless species that depend on them for survival. For centuries, understanding these aquatic ecosystems required researchers to wade into the water, relying on nets and human eyes to count fish and guess at their health. Today, a new approach is emerging that blends the physical world with digital intelligence. This method relies on the Internet of Things, a concept where everyday objects are connected to the internet to share information, and underwater sensor networks, which act like a nervous system beneath the surface. By placing smart devices in rivers, scientists can now gather data continuously without constant human presence. The goal is simple yet profound: to watch over fish populations with the same care a gardener gives to a garden, using technology to spot changes in real time and protect the delicate balance of life underwater.

In a recent study, researchers from institutions in Pakistan have built a system that brings this vision to life, specifically designed to monitor fish in river environments. They created a digital framework that uses cameras equipped with artificial intelligence to take pictures of fish and then instantly identifies what species are present. Imagine a camera that does not just see a shape in the water but understands it, recognizing a trout from a salmon without needing a human to look at the photo first. The team developed a software platform that runs on a standard desktop computer, acting as the brain of the operation. This software connects to the cameras, processes the images, and displays the results on a screen where anyone can see the current number of fish and their types. The system is designed to work automatically, taking pictures at set intervals and updating the counts without anyone needing to press a button, making it possible to watch a river's life unfold from a desk miles away.

The heart of this project is a piece of software trained to recognize fish, much like a person learns to identify birds by their feathers and shape. The researchers taught the computer using a large collection of photos showing different kinds of fish in various lighting conditions and angles. Once the computer learned these patterns, they tested it with new images. The results were encouraging; the system correctly identified the fish in the test images about ninety-two percent of the time. It successfully distinguished between species like horse mackerel, shrimp, red mullet, and trout, displaying the counts on a dashboard that updates as new data comes in. The team also built a secure login system so that only authorized people could view or manage the data, ensuring that the information remains safe while still being accessible to those who need it.

To make sure the system works as intended, the researchers ran a series of checks. They verified that the cameras could capture clear images, that the software could count the fish accurately, and that the numbers appeared on the screen immediately without delay. They also tested the ability to add new types of fish to the system's memory, allowing the tool to grow and adapt as more species are discovered or as the river's population changes. The entire setup relies on a local interface where users can select image files for processing, running on a desktop computer to simulate the monitoring process. While the system was tested in a simulated environment to prove the concept works, the researchers noted that it is currently limited to a digital simulation and cannot yet be implemented on surface water routing information.

The work represents a significant step forward in how we watch over our waterways, offering a way to gather information that was previously difficult or impossible to obtain. By automating the detection and counting of fish, the system frees up time for researchers to focus on what the data means rather than just collecting it. The author acknowledges that there is still work to be done, such as improving the system's ability to handle different underwater conditions and expanding the library of fish it can recognize. They also plan to test the technology in real-world settings, moving beyond the computer simulation to see how it performs in a flowing river. For now, this project stands as a proof of concept, showing that with the right mix of cameras, smart software, and internet connections, we can begin to listen to the quiet stories rivers tell about the life they hold.