Ultrasonic touch detection and decision

Abstract

A method includes receiving energy data associated with an ultrasound input device coupled to a material layer. The energy data comprises a current energy value and past energy values associated with reflected ultrasound signals received at the ultrasound input device in response to the ultrasound input device transmitting emitted signals through the material layer towards an external surface of the material layer. The method can then include comparing the energy data with threshold data to generate a current trigger value for trigger data. The trigger data is indicative of an occurrence of a touch event when the current energy value exceeds a current threshold value of the threshold data. Then the method can include updating the threshold data based on the energy data, the trigger data, and the threshold data. Updating the threshold data comprises generating a subsequent threshold value.

What is claimed is: 
     
         1 . A method, comprising:
 receiving energy data associated with an ultrasound input device coupled to a material layer, wherein the energy data comprises a current energy value and past energy values associated with reflected ultrasound signals received at the ultrasound input device in response to the ultrasound input device transmitting emitted signals through the material layer towards an external surface of the material layer;   comparing the energy data with threshold data to generate a current trigger value for trigger data, wherein the trigger data is indicative of an occurrence of a touch event when the current energy value exceeds a current threshold value of the threshold data; and   updating the threshold data based on the energy data, the trigger data, and the threshold data, wherein updating the threshold data comprises generating a subsequent threshold value.   
     
     
         2 . The method of  claim 1 , further comprising:
 receiving a subsequent value associated with the energy data; and   comparing the subsequent value associated with the energy data to a subsequent threshold signal to generate a subsequent trigger value associated with the trigger data.   
     
     
         3 . The method of  claim 1 , wherein updating the threshold data is based on the current energy value and a preset number of past energy values, and wherein the subsequent threshold value is greater than the current threshold value when the energy data has an upward trend. 
     
     
         4 . The method of  claim 1 , wherein the trigger data comprises the current trigger value and past trigger values, wherein updating the threshold data is based on the current trigger value and a preset number of the past trigger values. 
     
     
         5 . The method of  claim 1 , wherein updating the threshold data comprises determining a speed of change in value of the energy data over a duration and generating the subsequent threshold value based on the determined speed. 
     
     
         6 . The method of  claim 1 , wherein updating the threshold data comprises determining an extent of change in value of the energy data over a duration and generating the subsequent threshold value based on the determined extent. 
     
     
         7 . The method of  claim 1 , wherein updating the threshold data comprises determining a number of touch events indicated by the trigger data for a duration and generating the subsequent threshold value based on the number of touch events indicated by the trigger data for the duration. 
     
     
         8 . The method of  claim 1 , wherein comparing the energy data with the threshold data to generate the current trigger value further comprises using past trigger values of the trigger data to confirm that the touch event has occurred. 
     
     
         9 . The method of  claim 1 , wherein updating the threshold data based on the energy data, the trigger data, and the threshold data comprises applying an adaptive threshold update function, wherein the adaptive threshold update function is configured to permit slower changes in the energy data to have more effect on the threshold data than faster changes in the energy data. 
     
     
         10 . The method of  claim 1  further comprising:
 analyzing the trigger data to determine a sensor state and generating an output based on the sensor state. 
 
     
     
         11 . A method comprising:
 receiving energy data associated with an ultrasound input device coupled to a material layer, wherein the energy data comprises a current energy value and past energy values associated with reflected ultrasound signals received at the ultrasound input device in response to the ultrasound input device transmitting emitted signals through the material layer towards an external surface of the material layer; and   providing the energy data to a recurrent neural network to generate output data indicative of an occurrence of a touch event at the external surface of the material layer.   
     
     
         12 . The method of  claim 11  further comprising:
 determining a state classification associated with the touch event based on trigger values. 
 
     
     
         13 . The method of  claim 11 , wherein the recurrent neural network is trained using historical energy data associated with a plurality of historical touch events. 
     
     
         14 . The method of  claim 13 , wherein the plurality of historical touch events comprises one or more of each of a set of state classifications. 
     
     
         15 . The method of  claim 14 , wherein the set of state classifications is selected by user input out of a plurality of available state classifications. 
     
     
         16 . The method of  claim 14 , wherein the plurality of historical touch events further comprises a plurality of non-touch events to facilitate training an additional recurrent neural network to reject false positive events. 
     
     
         17 . The method of  claim 11 , wherein the output data comprises state classification information associated with the touch event. 
     
     
         18 . The method of  claim 11 , further comprising providing the output data from the recurrent neural network to an additional recurrent neural network to generate state classification information associated with the touch event. 
     
     
         19 . The method of  claim 18 , wherein the additional recurrent neural network is trained using historical energy data associated with a plurality of historical touch events at the external surface, and wherein the plurality of historical touch events comprises one or more of each of a set of state classifications. 
     
     
         20 . The method of  claim 19 , wherein the set of state classifications is selected by user input out of a plurality of available state classifications. 
     
     
         21 . The method of  claim 19 , wherein the plurality of historical touch events further comprises a plurality of non-touch events to facilitate training the additional recurrent neural network to reject false positive events. 
     
     
         22 . The method of  claim 11 , wherein the recurrent neural network comprises a first hidden layer and a second hidden layer, wherein a tapped delay line of the output data is provided as an input to the first hidden layer, and wherein outputs from the first hidden layer are provided as inputs to the second hidden layer. 
     
     
         23 . The method of  claim 11 , wherein providing the energy data to the recurrent neural network comprises providing current energy data and a preset number of past energy values to the recurrent neural network. 
     
     
         24 . A controller configured to perform:
 receiving energy data associated with an ultrasound input device coupled to a material layer, wherein the energy data comprises a current energy value and past energy values associated with reflected ultrasound signals received at the ultrasound input device in response to the ultrasound input device transmitting emitted signals through the material layer towards an external surface of the material layer;   comparing the energy data with threshold data to generate a current trigger value for trigger data, wherein the trigger data is indicative of an occurrence of a touch event when the current energy value exceeds a current threshold value of the threshold data; and   updating the threshold data based on the energy data, the trigger data, and the threshold data, wherein updating the threshold data comprises generating a subsequent threshold value.   
     
     
         25 . The method of  claim 24 , further comprising:
 receiving a subsequent value associated with the energy data; and   comparing the subsequent value associated with the energy data to a subsequent threshold signal to generate a subsequent trigger value associated with the trigger data.   
     
     
         26 . The method of  claim 24 , wherein updating the threshold data is based on the current energy value and a preset number of past energy values, and wherein the subsequent threshold value is greater than the current threshold value when the energy data has an upward trend. 
     
     
         27 . The method of  claim 24 , wherein the trigger data comprises the current trigger value and past trigger values, wherein updating the threshold data is based on the current trigger value and a preset number of the past trigger values. 
     
     
         28 . A controller configured to perform:
 receiving energy data associated with an ultrasound input device coupled to a material layer, wherein the energy data comprises a current energy value and past energy values associated with reflected ultrasound signals received at the ultrasound input device in response to the ultrasound input device transmitting emitted signals through the material layer towards an external surface of the material layer; and   providing the energy data to a recurrent neural network to generate output data indicative of an occurrence of a touch event at the external surface of the material layer.   
     
     
         29 . The method of  claim 28  further comprising:
 determining a state classification associated with the touch event based on trigger values. 
 
     
     
         30 . The method of  claim 28 , wherein the recurrent neural network is trained using historical energy data associated with a plurality of historical touch events.