International Journal of Scientific Research and Engineering Development

The artificial intelligence and computer vision are changing fast. This means we can think of ways for people to interact with computers. Usually, we use the keyboard and mouse to control the computer. These need to be touched which is a problem for people who have trouble moving their hands and arms and it is not clean in places where we need to be careful about germs. This paper is about a system that lets people control the computer cursor with their hand movements. We use a webcam to see what the persons hands are doing. The system uses OpenCV to look at the video from the webcam and get the information we need and MediaPipe to find the points on the persons hands in real time. Then we use a set of rules to figure out what the person wants to do with the computer based on how their fingers are moving. The system can do all the things a regular mouse can do like moving the cursor clicking the button and clicking the right button. We also have a way to make the cursor move smoothly on the screen even if the persons hand is shaking a bit. The best part is that we do not need any hardware, just a regular webcam. We tried out the system in lighting and backgrounds and it worked really well. The system was able to understand what the person was trying to do with their hands 95.8 percent of the time and it could move the cursor to the right place about 96.2 percent of the time. It could also tell when the person wanted to click the button 94.7 percent of the time and it only took about 0.15 seconds to respond. These results show that the system is good and can be used to help people who have trouble using computers and it can also be used in places where we need to be careful about germs. The artificial intelligence and computer vision are really helping to make this possible. We think it is a big step forward for accessibility applications and contactless computing environments, with the artificial intelligence and computer vision.

Shifa Bilal Tamboli, Simeen Phiroj Mulani, Arman Tajuddin Shiakh,"Deep Learning-Based Gesture Interface for Cursor Control Using Computer Vision" International Journal of Scientific Research and Engineering Development, V9(2): Page(2563-2570) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.