A computer vision and robotics expert has created and trained what he boasts is “the ultimate mosquito killer.” Steven Cheng shared details of his high-tech bug zapper project on social media. Key innovations here include the use of computer vision and deep learning technologies to detect and lock onto mosquitoes so the laser ‘artillery cannon’ could do its work.
Spent 4 months building the ultimate mosquito killer: an artillery cannon guided by computer vision + deep learning.Trained a custom model to detect and lock onto mosquitoes using a DSLR + zoom lens setup.The dataset collection phase was brutal — the mosquitoes definitely… pic.twitter.com/jqfgz0eq9lMay 28, 2026
For scanning the environment, Cheng concluded that a DSLR paired with a high-magnification zoom lens was the best option. This was also used in the training stage to build up a large dataset of mosquito images. A side effect of ‘welcoming’ mosquitoes in for photographs at this stage of the project was “countless mosquito bites all over my body,” recalled the intrepid technologist.
With the image database built and annotated, Cheng moved on to leveraging deep learning techniques. This task “really put my graphics card through its paces,” he commented. However, the detection performance of the resulting model was “quite good” by the end of this process.
A laser source was tuned to “instantly turn mosquitoes into roasted ones.” Then, the equipment was mounted on a high-precision industrial rotary stage / gimbal to complete the ‘ultimate mosquito killer’ apparatus.
Simulations were run, and Cheng decided to add a wide-angle camera to the setup. The purpose of the second camera with a wider view was to detect humans and flammable materials in the house. Logic was implemented where, if there was any overlap between humans or flammable materials and the laser target, no power would be fed to the cannon.
Overall, Cheng was pleased with the results and says all the mosquitos in his residence were “successfully eliminated” after a night’s effort.
Rival Indiegogo product design ships in June
If the above story sounds familiar, that’s because it isn’t the first laser-toting mosquito killer for the home that we’ve reported on. Last year we highlighted the Photonmatrix Indiegogo project, which sought funding for an all-in-one portable laser-driven mosquito-killing machine costing as little as $498.
The Photonmatrix was claimed to leverage a LiDAR scanner combined with a galvanometer-directed laser to seek and destroy mosquitoes at a rate of up to 30 pests per second. However, its detection method doesn’t sound as impressive as the machine-learning tech used by Cheng’s new device.
Hopefully we will see consumer reviews of the Photonmatrix soon, though, as it is due to ship very shortly. Backers will receive devices starting from “June 2026,” according to the latest information.
Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.
