# US 11,029,785 B2 · Method for improving accuracy of touch screen event analysis by use of spatiotemporal touch patterns > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,029,785 B2 - Document ID: US11029785B2 - Status: Granted - Kind: B2 - Application number: 16/798,139 - Filed: 2020-02-21 - Published: 2020-07-02 - Granted: 2021-06-08 - Earliest claimed: 2014-09-24 - Assignees: Qeexo, Co. - Applicants: Qeexo, Co. - Inventors: Julia Schwarz; Christopher Harrison - Web page: https://patentgrep.com/patent/US11029785B2 ## CPC - G06F 3/043: using propagating acoustic waves - G06F 3/0416: {Control or interface arrangements specially adapted for digitisers} - G06F 2203/04104: Multi-touch detection in digitiser, i.e. details about the simultaneous detection of a plurality of touching locations, e.g. multiple fingers or pen and finger - G06F 2203/04106: Multi-sensing digitiser, i.e. digitiser using at least two different sensing technologies simultaneously or alternatively, e.g. for detecting pen and finger, for saving power or for improving position detection - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract A method of classifying touch screen events uses known non-random patterns of touch events over short periods of time to increase the accuracy of analyzing such events. The method takes advantage of the fact that after one touch event, certain actions are more likely to follow than others. Thus if a touch event is classified as a knock, and then within 500 ms a new event in a similar location occurs, but the classification confidence is low (e.g., 60% nail, 40% knuckle), the classifier may add weight to the knuckle classification since this touch sequence is far more likely. Knowledge about the probabilities of follow-on touch events can be used to bias subsequent classification, adding weight to particular events. ## Claims ### Claim 1 A system for analyzing touch events based on characterization of features derived from the touch events, the system comprising: a touch sensitive surface for receiving touch events thereon; at least one sensor for generating vibro-acoustic waveform signals associated with the touch events; a touch event detector for detecting the touch events, including an initial touch event from an initial hand part and a subsequent touch event, on the touch sensitive surface; a classifier module for: converting the vibro-acoustic waveform signals into at least one domain signal; extracting distinguishing features from the at least one domain signal; and classifying the distinguishing features of the subsequent touch event by employing spatiotemporal data relating to historical patterns for different sequences of hand parts causing historical initial and subsequent touch events, wherein the classifying uses the spatiotemporal data to derive different likelihoods for different hand parts causing the subsequent touch event, which occurs after the initial hand part caused the initial touch event, to determine which subsequent hand part from the different hand parts caused the subsequent touch event, and wherein the different hand parts include a finger knuckle, a fingertip, a palm, and a fingernail. ### Claim 2 (depends on 1) The system recited in claim 1, wherein the historical patterns associated with the spatiotemporal data comprises data from prior sequence of multiple touch events over a selected time period. ### Claim 3 (depends on 1) The system recited in claim 1, wherein the spatiotemporal data comprises data suggesting a likelihood of each touch event after a first of a plurality of touch events over a selected time period. ### Claim 4 (depends on 1) The system recited in claim 1, wherein the historical patterns associated with the spatiotemporal data is accumulated and stored in advance of the classifying. ### Claim 5 (depends on 1) The system recited in claim 1, wherein the classifying the distinguishing features comprises multiplying spatiotemporal event likelihood by classification confidences. ### Claim 6 (depends on 1) The system of claim 1, further comprising an operating system for activating a function with respect to the touch sensitive surface, wherein such function is based on classifying the distinguishing features. ### Claim 7 A system for classifying a sequence of multiple touch events in touch screen devices to improve accuracy of determining characteristics of touch events imparted to a touch screen, the system comprising: a touch screen for receiving the touch events imparted thereon; a classifier module for: analyzing each touch event in said sequence to determine within a level of confidence what hand part caused a touch event; determining likelihoods of different sequences of possible hand part touch events based on spatiotemporal event data relating to historical patterns of different sequences of hand parts that cause the different sequences of touch events; and combining the level of confidence for each touch event in said sequence and the likelihoods of different sequences of the possible hand part touch events to produce a most likely result of what hand part actually generated the touch event in the sequence, wherein such hand part is selected from among a fingertip, a finger knuckle, a fingernail, and a hand palm. ### Claim 8 (depends on 7) The system recited in claim 7, wherein the determining uses elapsed time over the sequence of multiple touch events to determine the likelihoods. ### Claim 9 (depends on 7) The system recited in claim 7, wherein the combining comprises weighting the confidence levels by the likelihoods. ### Claim 10 (depends on 7) The system recited in claim 7, wherein the combining comprises multiplication of the confidence levels by the likelihoods. ### Claim 11 A method of analyzing touch events based on characterization of features derived from the touch events, the method comprising: detecting the touch events, including an initial touch event from an initial hand part and a subsequent touch event, on a touch sensitive surface; generating vibro-acoustic waveform signals using at least one sensor associated with the touch events; converting the vibro-acoustic waveform signals into at least one domain signal; extracting distinguishing features from the at least one domain signal; classifying the distinguishing features of the subsequent touch event by employing spatiotemporal data relating to historical patterns for different sequences of hand parts causing historical initial and subsequent touch events, wherein the classifying uses the spatiotemporal data to derive different likelihoods for different hand parts causing the subsequent touch event, which occurs after the initial hand part caused the initial touch event, to determine which subsequent hand part from the different hand parts caused the subsequent touch event, and wherein the different hand parts include a finger knuckle, a fingertip, a palm, and a fingernail; and activating a function with respect to the touch sensitive surface, wherein such function is based on classifying the distinguishing features. ### Claim 12 (depends on 11) The method of claim 11, wherein the historical patterns associated with the spatiotemporal data comprises data from prior sequence of multiple touch events over a selected time period. ### Claim 13 (depends on 11) The method of claim 11, wherein the spatiotemporal data comprises data suggesting a likelihood of each touch event after a first of a plurality of touch events over a selected time period. ### Claim 14 (depends on 11) The method of claim 11, wherein the historical patterns associated with the spatiotemporal data is accumulated and stored in advance of the classifying. ### Claim 15 (depends on 11) The method of claim 11, wherein the classifying the distinguishing features comprises multiplying spatiotemporal event likelihood by classification confidences.