Skip to content

Gesture Recognition

Jump to bottom
Barryblueice edited this page Dec 31, 2025 · 1 revision

In Windows system, gesture recognition is automatically recognized by system base level, instead of directly "declared" by applications or touchpad hardware.

So the precision touchpad will only report the original touch data:

  • How many finger(s) are touching right now
  • Each finger(s)：
    • Unique ID
    • X/Y coordinates
    • Whether finger(s) to touch
    • Whether finger(s) to leave
  • Continuous multi-frame data over time

For example, if I do a three fingers touching, logically showing is:

Frame 1：3 points，location A/B/C
Frame 2：3 points，overall upward movement
Frame 3：3 points，keep moving upwards
Frame 4：points losing (fingers have left)

When original touch data is posted to system, Windows Precision ToucPad driver will:

  1. Analyze HID report
  2. Transfer each points into internal pointer objects
  3. Send these pointer data to the Input stack

The system will know

  • The numbers of n pointers exist at the same time
  • Their movement trajectories and life cycles

Finally, the system will determine the type of multi-finger touch through a non-public gesture recognition engine.

Clone this wiki locally