Multi-dimensional multi-finger search using oversampling hill climbing and descent with range

• US 10,073,552 B2
• Filed: 01/15/2014
• Issued: 09/11/2018
• Est. Priority Date: 01/15/2013
• Status: Active Grant

• Current Assignee: Cirque Corporation
• Original Assignee: Cirque Corporation
• Inventors: Layton, Michael D.
• Primary Examiner(s): Hicks, Charles
• Application Number:
  US14/156,087
• Publication Number:
  US 20140198076A1
• Time in Patent Office: 1,700 Days
• CPC Class Codes: 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
• CPC Class Codes: G06F 3/04166 Details of scanning methods, e.g. sampling time, grouping of sub areas or time sharing with display driving Synchronisation with the driving of the display or the backlighting unit to avoid interferences generated internally G06F3/04184

Abstract

A method for detecting and tracking multiple fingers in a two dimensional set of results using oversampling hill climbing and descent with range.

Claims

1. A method for detecting and tracking multiple objects using a two dimensional set of measurements from a touch sensor, said method comprising:

1) scanning a two-dimensional array of locations to identify a first finger location having a result which is greater than zero such that the first finger location indicates the presence of at least one finger;

2) performing a hill climbing algorithm in a measurement cycle to determine if there is a location immediately adjacent to the first finger location that has a higher value which does not cause the hill climbing procedure to enter into an infinite loop around an actual location of a finger that causes the hill climbing algorithm to not identify the first finger location with a maximum result, and terminating the infinite loop when the hill climbing procedure repeats a location in the current measurement cycle;

3) repeating the hill climbing algorithm until the highest first finger location is found; and

4) performing a hill descending algorithm to identify a boundary between the first finger location and a second finger.

2. The method as defined in claim 1 wherein the method further comprises oversampling the two dimensional array of locations when scanning to thereby make certain that at least one adjacent location to the first finger location also has a result to thereby confirm the presence of at least one finger.

3. The method as defined in claim 1 wherein the hill climbing algorithm further comprises bypassing local minima caused by aliasing in order to find the first finger location with the highest value to create a new current first finger location by comparing a weighted sum of near cross axis results.

4. The method as defined in claim 3 wherein the hill climbing algorithm further comprises performing a 4-connected comparison of the four closest locations in order to determine the new current first finger location.

5. The method as defined in claim 4 wherein comparing a weighted sum of near cross axis results further comprises creating separate filtered data arrays for each climbing direction around the current first finger location.

6. The method as defined in claim 5 wherein the weighted sum further comprises applying a weighted pattern that gives more weight to some locations and less weight to other locations.

7. The method as defined in claim 1 wherein the hill descending algorithm further comprises including a range of results which are then considered to be part of the same finger.

8. The method as defined in claim 7 wherein the hill descending algorithm further comprises identifying a boundary with an adjacent finger by comparison of adjacent results, with range.

9. The method as defined in claim 8 wherein the method further comprises repeating steps 1 through 4 while excluding any fingers that are already identified.

17. The method as defined in claim 1 wherein the hill climbing algorithm further comprises terminating the infinite loop when the hill climbing algorithm has moved N number of times, where N is selected to be the number of steps that are sufficient to determine that the hill climbing algorithm has found the maximum result.

10. A method for detecting and tracking a single object using a two-dimensional set of measurements from a touch sensor, said method comprising:

1) scanning a two-dimensional array of locations to identify a finger location having a result which is greater than zero such that the first finger location indicates the presence of the finger;

2) performing a hill climbing algorithm in a measurement cycle to determine if there is a location immediately adjacent to the finger location that has a higher value which does not cause the hill climbing procedure to enter into an infinite loop around an actual location of a finger that causes the hill climbing algorithm to not identify the first finger location with a maximum result, and terminating the infinite loop when the hill climbing procedure repeats a location in the current measurement cycle; and

3) repeating the hill climbing procedure until the highest finger location is found.

11. The method as defined in claim 10 wherein the method further comprises oversampling the two dimensional array of locations when scanning to thereby make certain that at least one adjacent location to the finger location also has a result to thereby confirm the presence of the finger.

12. The method as defined in claim 10 wherein the hill climbing algorithm further comprises bypassing local minima caused by aliasing in order to find the finger location with the highest value to create a new finger location by comparing a weighted sum of near cross axis results.

13. The method as defined in claim 12 wherein the hill climbing algorithm further comprises performing a 4-connected comparison of the four closest locations in order to determine the new current finger location.

14. The method as defined in claim 13 wherein comparing a weighted sum of near cross axis results further comprises creating separate filtered data arrays for each climbing direction around the current finger location.

15. The method as defined in claim 14 wherein the weighted sum further comprises applying a weighted pattern that gives more weight to some locations and less weight to other locations.

16. The method as defined in claim 10 wherein the hill descending algorithm further comprises including a range of results which are then considered to be part of the finger.

18. The method as defined in claim 10 wherein the hill climbing algorithm further comprises terminating the infinite loop when the hill climbing algorithm has moved N number of times, where N is selected to be the number of steps that are sufficient to determine that the hill climbing algorithm has found the maximum result.