GB2544258A

GB2544258A - Gesture detection mechanism

Info

Publication number: GB2544258A
Application number: GB1516632.5A
Authority: GB
Inventors: Philip Pisani Justin; Robert Gordon Simon
Original assignee: Individual
Current assignee: Individual
Priority date: 2015-09-19
Filing date: 2015-09-19
Publication date: 2017-05-17
Also published as: GB201516632D0

Abstract

A personal sensor device equipped with an accelerometer 2, a microprocessor 4 to record accelerations and recognise one or more arm movement gestures and being able to adapt the thresholds for detecting the gesture depending on the variation in acceleration due to activity. The method to adapt the threshold may be linear, logarithmic or a combination of the two and may be learned during usage of the sensor device. An initial device orientation may be used to reduce false gesture detection at lower levels of activity. The device may comprise a display 1 or use another method of user feedback that indicates to the user when the device is in an appropriate state to accept a gesture triggered action. The device may include a vibrator 6 that indicates to the user when they have made a movement that is a valid part of a gesture.

Description

Gesture detection mechanism

Description

Wearable personal computing devices which have access to the fixed and mobile internet are now commonplace. Often these devices use microminiature accelerometers (e.g. ST Microelectronics LIS2DH) to detect accelerative forces when worn by a user.

Commonly these forces can be used to detect physical actions such as waving or shaking and use these actions to trigger device functions.

In current devices a problem exists with the reliability of distinguishing intentional gestures from unintended gestures, particularly when allowing for differences in a user's activities of daily living at different times.

This invention proposes a method for the personal sensing device to adapt how it determines accelerative force gestures in order to minimise false detections.

Preferred Embodiment

The invention is now described with reference to the following figures:

Figure 1 shows an example personal worn sensor device Figure 2 shows a method to improve the detection mechanism.

Referring to Figure 1, the personal sensor device is worn on the wrist or carried in the hand by the user as they undertake activities of daily living and contains a microprocessor [4] and wireless transceiver [5]. Those skilled in the art will be aware that these functions may be achieved using a single electronic component such as wireless System on Chip (Texas Instruments, USA). The device measures acceleration forces using an accelerometer [2] as well as other environmental parameters such as temperature. The device may be able to interpret these parameters and display them to the user via a user interface [1], The device may also provide haptic feedback using a vibrator [6], The unit may be powered by a battery [5] such as a generic CR2032 coin cell (Duracell Inc, USA).

The accelerometer [2] may provide a measure of acceleration in 3 or 6 axes. The sensor may determine a gesture, such as a horizontal shake, by monitoring the accelerative change of the sensor using the microprocessor [4] and determining if a threshold has been exceeded.

The accelerometer may also be used to determine the orientation of the device by measuring the accelerative force on each device axis due to gravity.

The reliability of the sensed gesture may be improved by combining a starting orientation followed by an acceleration change in an axis to define a specific gesture.

Under motion of walking or running, the forces generated by the user's gait will mean that the sensing threshold values which are optimal at rest become suboptimal. Referring to figure 2, the invention uses the method described to maintain optimal threshold values.

On start [7] the sensor sets its gesture thresholds [8] to levels suitable for operation when the user is static, consistent with comfortable physical arm movement with a low level of force. As noted earlier this can be combined with a defined starting orientation to reduce accidental triggering due to normal arm movements.

The microprocessor [4] then moves via transition [9] to monitoring movement accelerations and checking to see if these are changing [10].

If an increase in activity is detected [12] via transition [11], the microprocessor adjusts the gesture thresholds [14] in relation to the activity increase detected at [10], Those skilled in the art will be aware that the algorithm to adjust the thresholds may be linear or logarithmic in nature or a combination of both. The microprocessor then transitions via [19] back to monitoring movement accelerations [10].

If a decrease in activity is detected [16] via transition [15], the microprocessor adjusts the gesture thresholds [18] in relation to the activity decrease detected at [10] using a corresponding algorithm to [14], The microprocessor then transitions via [20] back to monitoring movement accelerations [10].

Claims

Claims We claim:-

1. A personal sensor device equipped with: a. an accelerometer b. a microprocessor means to record accelerations and recognise one or more arm movement gestures c. a method to adapt the thresholds for detecting the gesture depending on the variation in acceleration due to activity

2. A system as defined in claim 1, where the method to adapt the threshold may be linear, logarithmic or a combination of the two.

3. A system as defined claim 1, where the method to adapt the threshold may be learned during usage of the sensor device.

4. A system as defined in claim 1, where an initial device orientation may be used to reduce false gesture detection at lower levels of activity.

5. A system as defined in claim 1 with a display or other method of user feedback that indicates to the user when the device is in an appropriate state to accept a gesture triggered action.

6. A system as defined in claim 1 with vibrator that indicates to the user when they have made a movement that is a valid part of a gesture.