Signal adaptive filtering for touch detection

Abstract

A signal adaptive filtering technique for recognizing touch and release events as indicated from a measured capacitance signal received from a capacitive touch sensor device in order to improve electromagnetic noise immunity, event detection responses, adaptability to dynamically changing environments, and adaptability to signal sensitivity changes and signal offset over an extended period of time. A capacitive touch sensor system may include one or more capacitive touch sensor devices, each sending a measured capacitance signal that outputs a baseline capacitance signal during a release event, and outputs an increased capacitance signal during a touch event. The system updates the baseline capacitance signal with a filter, wherein a transfer function of the filter varies as a function of a current state of the capacitive touch sensor device, determines a range of capacitance values calculated from the baseline capacitance signal in which a recognition of the touch event will not be performed, creates a predicted signal from a filtered version of the measured capacitance signal, and recognizes a touch or release event from a comparison of the measured capacitance signal to the predicted signal.