~~ Actually, I made it long before the rise of Telewa. .. ~~
Introduction
Wouldn't it be nice if you could turn the screen when you're eating on the telewa and browsing the Web ** when both hands are closed ** ?? It's a burden on your wrist to bring your hand to the mouse every time. While leaning on the chair, if you raise your finger, you can switch tabs ...
Abstract
In summary, it's ** "Operation with shadow play" **.
Annotate the image you want to act on first.
Next, define the key input by mouse operation linked to the image in the config file. Then run the program and if it meets the OpenCV image similarity threshold, it will operate according to the definition.
I'm here, so please look at the repository for details.
As long as there is a shape, it can be replaced with ** whatever ** operation.
Is it okay to enter yoga poses and karate kata as a measure against lack of exercise? ~~ The work will never end! !! ~~
Hmm? Rather, is it an implementation suitable for people with some kind of physical handicap? ??
~~ I really wanted to implement it with tensorflow, but there is no GPU distributed by SIer, and real-time recognition is too far for PC. .. ~~
If you want to keep your hands on the keyboard, buy a 100% mirror as shown below and support analog! Lol
FYI
Similarity detection with Python ③ Feature point matching
Recommended Posts