How Touchscreens Use Math to Recognize Your Fingers

Apple's patent on using mathematical ellipses to track and identify individual fingers and palms on a touch-sensitive surface.

What it covers

This patent describes a method for a touchscreen to interpret raw electrical data as distinct physical objects. It takes a 'proximity image'—a grid of data from electrodes—and groups nearby pixels together to represent a touch. The system then mathematically fits an ellipse to these pixel groups to estimate the shape, size, and orientation of the finger or palm. By tracking these ellipses over time, the device can distinguish between a thumb and a fingertip or detect if a hand is resting on the screen.

What it doesn't cover

• —Does not cover touch detection methods that do not use ellipse fitting (e.g., simple centroid or bounding box tracking).
• —Does not cover hardware that lacks a proximity-sensing electrode grid.
• —Does not cover software gestures that are not derived from the calculated ellipse parameters.

The clever bit

Instead of treating a touch as a single point (an X,Y coordinate), the inventors treated it as a shape with orientation, using eigenvectors and eigenvalues to calculate the ellipse that best fits the contact area.

Why it matters

This technology was foundational for the transition from simple single-touch screens to the sophisticated multi-touch interfaces found in modern smartphones and trackpads. By allowing the software to 'understand' the difference between a finger and a palm, it enabled features like palm rejection, which prevents accidental clicks while typing.

Real-world examples

• 1.Apple MacBook trackpads
• 2.iPhone and iPad multi-touch displays
• 3.Modern laptop touchscreens with palm rejection