Method and device for the recognition of the position or movement of a device or a person

• US 8,301,575 B2
• Filed: 07/23/2008
• Issued: 10/30/2012
• Est. Priority Date: 07/23/2007
• Status: Active Grant

• Current Assignee: Commissariat a L'Energie Atomique (AES Corporation)
• Original Assignee: Commissariat a L'Energie Atomique (AES Corporation)
• Inventors: Bonnet, Stephane, Godin, Christelle
• Primary Examiner(s): Gaffin, Jeffrey A
• Assistant Examiner(s): Kim, David H
• Application Number:
  US12/178,559
• Publication Number:
  US 20090030345A1
• Time in Patent Office: 1,560 Days
• Field of Search: 706/12
• US Class Current: 706/12
• CPC Class Codes: A61B 2560/0418 Pen-shaped housings
• CPC Class Codes: A61B 2562/0219 Inertial sensors, e.g. accelerometers, gyroscopes, tilt switches
• CPC Class Codes: A61B 5/1116 Determining posture transitions
• CPC Class Codes: A61B 5/1123 Discriminating type of movement, e.g. walking or running A61B5/1116, A61B5/112 take precedence
• CPC Class Codes: A61B 5/4094 Diagnosing or monitoring seizure diseases, e.g. epilepsy
• CPC Class Codes: A61B 5/4528 Joints A61B5/4533, A61B5/4538 take precedence
• CPC Class Codes: A61B 5/6814 Head
• CPC Class Codes: A61B 5/6823 Trunk, e.g., chest, back, abdomen, hip
• CPC Class Codes: A61B 5/6824 Arm or wrist
• CPC Class Codes: A61B 5/6828 Leg
• CPC Class Codes: A61B 5/6898 Portable consumer electronic devices, e.g. music players, telephones, tablet computers
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: G06V 30/142 using hand-held instruments; Constructional details of the instruments
• CPC Class Codes: G06V 30/242 Division of the character sequences into groups prior to recognition; Selection of dictionaries
• CPC Class Codes: G06V 40/23 Recognition of whole body movements, e.g. for sport training
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

This method uses two sets of sensors to estimate certain characteristics of the movement of a device or a person or states, especially postures, they adopt. A first, abundant, set of sensors (1) is removed after a learning phase where it records with certainty the states obtained by interpreting first decisional rules. The measurements of a second set of sensors (2), much more restricted than the first, are correlated to the states reached during the learning period by second decisional rules automatically obtained by supplying a classifier. They are then interpreted to determine the new states reached by the wearer just by means of the second sensors. The results are good in spite of the low number of second sensors, thanks to the accuracy of the second decisional rules.

Claims

1. Method for the recognition of the position or movement of a device or a person, comprising the following steps:

providing first and second sets of sensors (1, 2, 5, 6, 12, 13, 14, 15) on or close to the device or person;

measuring, during a learning phase, the readings recorded by the sensors and applying first predefined decisional rules at least to the measurements taken by the first set of sensors to classify each event identified using the measurements of all or part of the sensors into a position or movement category;

creating a learning base, during the learning phase, identifying for all or part of said events their pre-established classification and the characteristics from the measurements of the second set of sensors;

defining second decisional rules based on the learning base permitting an event to be classified based on measurements taken by the second set of sensors;

removing the first set of sensors (1, 6, 12, 13) and only keeping the second set of sensors (2, 5, 14, 15) on the device or the person; and

classifying the position or the movement of the device or person, in a recognition phase, by applying the second decisional rules to the measurements made by the second set of sensors.

2. Method according to claim 1, characterised in that it is applied to the recognition of postures and movements, wherein the first and second sets of sensors comprise movement sensors such as accelerometers, magnetometers and gyrometers.

3. Method according to claim 2, characterised in that the second set of sensors comprises a tri-axial magnetometer and a tri-axial m accelerometer fitted to a wearer.

4. Method according to claim 3, characterised in that the first set of sensors comprises accelerometers fitted to the thighs and legs of the wearer.

5. Method according to claim 2, characterised in that the classified events comprise standing, seated and lying down postures as well as transitions between said standing, seated and lying down postures.

6. Method according to claim 1, characterised in that it is applied to the recognition of epileptic fits, wherein the first set of sensors comprises a sensor for the electrical activity of the brain of a person wearing the sensors, fitted to the wearer comprises movement sensors such as accelerometers, magnetometers and gyrometers.

7. Method according to claim 1, characterised in that it is applied to the recognition of epileptic fits, wherein the first set of sensors comprises an image recording device, and the second set of sensors fitted to the wearer comprises movement sensors such as accelerometers, magnetometers and gyrometers.

8. Method according to claim 1, characterised in that it is applied to the recognition of handwritten characters, wherein the first set of sensors is a graphic pad connected to sign identification means and the second set of sensors comprises movement sensors mounted on a pen.

9. Method according to claim 1, characterised in that it comprises driving a neurone system with the learning base to draft the second decisional rules.

10. Device for the application of a method for the recognition of a position or movement of an object or a person, comprising following steps:

providing first and second sets of sensors (1, 2, 5, 6, 12, 13, 14, 15) on or close to the object or person;

measuring, during a learning phase, the readings recorded by the sensors and applying first predefined decisional rules at least to the measurements taken by the first set of sensors to classify each event identified using the measurements of all or part of the sensors into a position or movement category;

creating a learning base, during the learning phase, identifying for all or part of said events their pre-established classification and the characteristics from the measurements of the second set of sensors;

defining second decisional rules based on the learning base permitting an event to be classified based on measurements taken by the second set of sensors;

removing the first set of sensors (1, 6, 12, 13) and only keeping the second set of sensors (2, 5, 14, 15) on the object or the person; and

classifying the position or the movement of the object or person, in a recognition phase, by applying the second decisional rules to the measurements made by the second set of sensors,

characterised in that it comprises the first set of sensors and the second set of sensors, and means of processing the measurements provided by the sensors containing a learning base and an adjustable classifier.