An engineer has created a laser system with AI to automatically fight mosquitoes

While most people escape mosquitoes with repellants, fumigators and mosquito nets, computer vision and robotics expert Steven Cheng decided to take a radical approach to the problem. An enthusiast has developed an experimental system that uses artificial intelligence, computer vision and a laser to automatically detect and destroy insects.

According to the author of the project, the goal was to create a completely autonomous installation capable of independently finding mosquitoes indoors and instantly neutralizing them without human intervention.

How the laser turret works

The basis of the system is a complex of computer vision, which analyzes the space in real time and tracks the movement of insects.

To find targets, Cheng used a digital SLR camera with a powerful telephoto lens. Such a decision made it possible to scan the premises efficiently and obtain images of sufficient quality for training the neural network.

At the initial stage, the developer had to collect his own database of mosquito photos. According to him, the process turned out to be quite painful: for the sake of creating a training dataset, he had to spend a lot of time with insects, which resulted in numerous bites.

The neuronet was taught to recognize mosquitoes

After the formation of the image base, work on deep learning algorithms began.

Cheng taught the model to recognize mosquitoes among other objects and track their position in space. Processing a large volume of data turned out to be a serious load for the graphics subsystem of his computer, but the final result, according to the developer, exceeded expectations.

The neural network could quite accurately determine the location of insects and transmit the coordinates to the laser guidance system almost in real time.

Security turned out to be one of the main challenges

To destroy mosquitoes, a laser is used, the power of which is sufficient to instantly destroy the target. The entire structure is mounted on a high-precision industrial rotary mechanism with a cardan guidance system, which allows you to quickly change the direction of shooting.

However, the developer paid special attention to security. Before launch, he conducted a series of computer simulations and tests to eliminate the risk of the laser accidentally affecting people or objects in the room.

In addition, the system received a wide-angle camera that constantly monitors the surrounding environment. If people, pets or potentially flammable materials appear in the area of ​​operation, the laser is automatically blocked until the risk is eliminated.

Test results

According to Cheng, during testing, the system showed high efficiency. The developer claims that in one night the installation was able to completely rid his house of mosquitoes.

Although the project is still experimental and not intended for mass production, it demonstrates how far modern computer vision and autonomous control technologies have advanced.