Continuous circle gesture detection for a sensor system
A method for detecting a continuous circle gesture, has the following steps: receiving vectors representative of an object movement by a object detection unit; determining from the received sequence velocity vectors a sequence of velocity vectors or an approximation thereof; estimating an angle between subsequent velocity vectors; and determining a rotation direction.
1. A method for detecting a continuous circle gesture, comprising
receiving a sequence of vectors representative of an object movement by a object detection unit, wherein the received sequence of vectors comprises measurement values m k (i) of electrodes i at times k;
determining from the received sequence of vectors a sequence of velocity vectors or an approximation thereof;
estimating an angle between subsequent velocity vectors; and
determining a rotation direction, wherein the rotation direction is determined by a sign of the angle.
2. The method according to claim 1 , wherein the received sequence of vectors (x n , y n ) represent position vectors of an object movement.
3. The method according to claim 2 , wherein velocity vectors are computed as the difference of subsequent position vectors.
4. The method according to claim 1 , wherein four measurement electrodes are provided and the velocity vectors v k are determined by
v
k
≈
[
(
m
k
(
4
)
-
m
k
-
1
(
4
)
)
-
(
m
k
(
2
)
-
m
k
-
1
(
2
)
)
(
m
k
(
3
)
-
m
k
-
1
(
3
)
)
-
(
m
k
(
1
)
-
m
k
-
1
(
1
)
)
]
.
5. The method according to claim 1 , wherein the angle between two subsequent velocity vectors
v
new
=
[
v
new
,
x
v
new
,
y
]
and
v
old
=
[
v
old
,
x
v
old
,
y
]
is determined by φ=arc cos( v new T · v old )·S, where Sε{±1} is the rotation direction, T indicates the vector transpose, and the bar indicates normalization of a vector.
6. The method according to claim 5 , wherein S:=sig n( v new,x v old,y − v new,y v old,x ).
7. The method according to claim 1 , wherein the angle φ between two subsequent velocity vectors v new and v old is approximately determined by φ≈∥ v new − v old ∥·S, where Sε{±1} is the rotation direction.
8. The method according to claim 7 , wherein S:=sig n( v new,x v old,y − v new,y v old,x ).
9. The method according to claim 1 , wherein the angle φ between two subsequent velocity vectors v new and v old is approximately determined by φ≈sin φ= v new,x v old,y − v new,y v old,x .
10. The method according to claim 1 , wherein velocity vectors v new =v k and v old =v k-1 .
11. The method according to claim 1 , wherein each velocity vector v new and v old in a sequence of two vectors is normalized by the same value ∥v new ∥ or ∥v old ∥.
12. The method according to claim 1 , wherein a circle counter is realized by integrating the angles between pairs of subsequent velocity vectors, or approximations of these angles.
13. The method according to claim 12 , wherein the circle counter is updated only if ∥v k ∥ exceeds a predetermined threshold.
14. The method according to claim 12 , wherein the received sequence and/or the velocity vectors and/or the circle counters are low-pass filtered.
15. The method according to claim 1 , wherein the sensor system is a 3-dimensional near-field sensor system or a 3-dimensional mid/far field sensor system.
16. The method according to claim 15 , wherein the near-field sensor system is a capacitive non-touching sensor system based on a quasi-static electric field measurements.
17. The method according to claim 15 , wherein the near-field capacitive sensor system comprises one or more transmission electrodes excited with a rectangular pulse train signal and a plurality of receiving electrodes capacitively coupled with the one or more transmission electrodes.
18. A human device interface comprising:
an interface generating a sequence of vectors representative of an object movement tracked by a gesture detection system, wherein the received sequence of vectors comprises measurement values m k (i) of electrodes i at times k;
a processing unit configured
to determine from the received vectors a sequence of velocity vectors or an approximation thereof;
to estimate an angle between subsequent velocity vectors; and
to determine a rotation direction, wherein the rotation direction is determined by a sign of the angle;
wherein the processing unit further implements a circle counter and updates values of the circle counter by adding the estimated angle with its sign depending on the rotating direction.
19. The human device interface according to claim 18 , wherein the received sequence of vectors (x n , y n ) represents position vectors of an object movement.
20. The human device interface according to claim 19 , wherein velocity vectors are computed as the difference of subsequent position vectors.
21. The human device interface according to claim 18 , wherein four measurement electrodes are provided and the velocity vectors v k are determined by
v
k
≈
[
(
m
k
(
4
)
-
m
k
-
1
(
4
)
)
-
(
m
k
(
2
)
-
m
k
-
1
(
2
)
)
(
m
k
(
3
)
-
m
k
-
1
(
3
)
)
-
(
m
k
(
1
)
-
m
k
-
1
(
1
)
)
]
.
22. The human device interface according to claim 18 , wherein the angle between two subsequent velocity vectors
v
new
=
[
v
new
,
x
v
new
,
y
]
and
v
old
=
[
v
old
,
x
v
old
,
y
]
is determined by φ=arc cos( v new T · v old )·S, where Sε{±1} is the rotating direction, T indicates the vector transpose, and the bar indicates normalization of a vector.
23. The human device interface according to claim 22 , wherein S:=sig n( v new,x v old,y − v new,y v old,x ).
24. The human device interface according to claim 18 , wherein the angle φ between two subsequent velocity vectors v new and v old is approximately determined by φ≈∥ v new − v old ∥·S, where Sε{±1} is the rotating direction.
25. The human device interface according to claim 24 , wherein S:=sig n( v new,x v old,y − v new,y v old,x ).
26. The human device interface according to claim 18 , wherein the angle φ between two subsequent velocity vectors v new and v old is approximately determined by φ≈sin φ= v new,x v old,y − v new,y v old,x .
27. The human device interface according to claim 18 , wherein velocity v vectors v new =v k and v old =v k-1 .
28. The human device interface according to claim 18 , wherein each velocity vector v new and v old in a sequence of two vectors is normalized by the same value ∥v new ∥ or ∥v old ∥.
29. The human device interface according to claim 18 , wherein a circle counter is realized by integrating the angles between pairs of subsequent velocity vectors, or approximations of these angles.
30. The human device interface according to claim 29 , wherein the circle counter is updated only if ∥v k ∥| exceeds a predetermined threshold.
31. The human device interface according to claim 29 , further comprising a first low-pass filter configured to filter the received sequence and/or a second low-pass filter configured to filter the velocity vectors and/or a third low-pass filter configured to filter the circle counter.
32. The human device interface according to claim 18 , wherein the interface comprises a 3-dimensional near-field sensor system, or a 3-dimensional mid/far field sensor system.
33. The human device interface according to claim 32 , wherein the near-field sensor system is a capacitive non-touching sensor system based on a quasi static electric field measurements.
34. The human device interface according to claim 33 , wherein the near-field capacitive sensor system comprises one or more transmission electrodes excited with a rectangular pulse train signal and a plurality of receiving electrodes capacitively coupled with the one or more transmission electrodes.
35. An electronic device comprising a human device interface according to claim 18 , wherein the circle counter is used for volume control, a dimmer, speed control, air condition temperature, or mechanical movement functions.
36. The electronic device according to claim 35 , wherein the circle counter drives an LED bar.
37. The method according to claim 1 , wherein when the angle φ>0, then the rotating direction is determined as a right rotation and when the angle φ<0, then the rotation direction is determined as a left rotation, wherein an angle of interest is within a range defined by
φ
=
⌈
-
π
2
,
π
2
⌉
.
38. The electronic device according to claim 18 , wherein when the angle φ>0, then the rotating direction is determined as a right rotation and when the angle φ<0, then the rotation direction is determined as a left rotation, wherein an angle of interest is within a range defined by
φ
=
⌈
-
π
2
,
π
2
⌉
.