Hand detection is a challenge topic in computer vision field. Many methods have been applied to try to detect hands precisely and real-time, however, the outcome looks not satisfying us. There still some way to go. Here i collect 10 inspiring videos to show what other people are doing to detect hands effectively.
This is an amazing demo that hands structures can be clearly re-modeled, although sometimes the detection looks a bit of flash and unstable. With the rebuilt structure of finger model, the next step of recognizing gesture would be easier based on such stable demo.
This is an old video of hands interaction demonstration. The author uses quite stable lib to detect hands’ gestures and get interactive manipulation over the recognization. It is stable, smart but it still can do better if improving the interaction responding time.
Not surprise, using the famous lib HandVu indeed can archive this effect. However, i tried to integrate the HandVu lib into my own project, but i failed due to the too far complexity of this lib.
This is just a video showing how hands can be located in the image and the direction of hands can be detected. It is quite practical, however, what if the pure background were changed to complicated one?
Ok, this is only a school project with purpose of testing the potential of hand detecting. Looks it is an preliminary result. Have fun.
2 Comments On “ 5 inspiring hands detection videos”