IMAGE CLASSIFICATION SYSTEM AND METHOD THEREOF

• US 20110033084A1
• Filed: 08/06/2009
• Published: 02/10/2011
• Est. Priority Date: 08/06/2009
• Status: Active Grant

• Current Assignee: Aptiv Technologies AG
• Original Assignee: Delphi Technologies, Inc. (BorgWarner Incorporated)
• Inventors: HE, XINHUA (SAM), NICHOLAS, CLAYTON L.
• Granted Patent: US 8,363,957 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06V 10/145 Illumination specially adapted for pattern recognition, e.g. using gratings
• CPC Class Codes: G06V 40/16 Human faces, e.g. facial parts, sketches or expressions

Abstract

An image classification system configured to classify a target and method thereof is provided, wherein the system includes at least one light source configured to emit light with at least one line pattern towards the target, wherein at least a portion of the emitted light and line pattern is reflected by the target. The system further includes an imager configured to receive at least a portion of the reflected light and line pattern, such that an obtained 2-D line pattern is produced that is representative of at least a portion of the emitted light and line pattern reflected by the target, and a controller configured to compare the 2-D line pattern to at least one previously obtained 2-D line pattern stored in a database, such that the controller classifies the 2-D line pattern as a function of the comparison.

Claims

1. An image classification system configured to classify a target comprising:

at least one light source configured to emit light towards the target, wherein at least a portion of said emitted light is reflected by the target;

an imager in optical communication with said at least one light source, wherein said imager is configured to receive at least a portion of said reflected light, such that a two-dimensional (2-D) line pattern is produced that is representative of at least a portion of said emitted light reflected by the target;

a database configured to store at least one previously obtained 2-D line pattern; and

a controller communicatively connected to said imager, wherein said controller is configured to compare said 2-D line pattern to said at least one previously obtained 2-D line pattern stored in said database, such that said controller classifies said 2-D line pattern as a function of said comparison.

2. The system of claim 1, wherein said at least one light source projects an initial line pattern at the target, such that the target is at least a portion of a face of a user of the image classification system.

3. The system of claim 1, wherein said obtained 2-D line pattern is a facial fingerprint, such that the target is at least a portion of a face of a user of the image classification system.

4. The system of claim 3, wherein said classification of said facial fingerprint is one of a full-registered user, a partial-registered user, and a non-registered user.

5. The system of claim 3, wherein said classification of said facial fingerprint is one of a forward facing classification, a rightward facing classification, leftward facing classification, an upward classification, a downward classification, an attentive classification, and a non-attentive classification.

6. The system of claim 1, wherein said 2-D line pattern has parameters of a 2-D line pattern image.

7. The system of claim 1, wherein said at least one light source is a plurality of light sources, and at least a portion of said plurality of light sources is configured to emit light at different wavelengths.

8. The system of claim 1 further comprising a 2-D line pattern generator in communicative connection with said imager, wherein said imager is configured to obtain an image of the target, and said 2-D line pattern generator is configured to produce said obtained 2-D line pattern as a function of said obtained image.

9. The system of claim 8, wherein said obtained 2-D line pattern is less data bytes than said image obtained by said imager of the target.

10. The system of claim 1 being used with a vehicle, such that the target is an occupant of said vehicle.

11. An image classification system configured to classify an awareness of an occupant of a vehicle, said system comprising:

at least one light source configured to emit light towards the occupant, wherein at least a portion of said emitted light is reflected by the occupant;

an imager in optical communication with said at least one light source, wherein said imager is configured to receive at least a portion of said reflected light, such that a two-dimensional (2-D) facial fingerprint is produced that is representative of at least a portion of said emitted light reflected by the occupant such that the 2-D facial fingerprint has parameters of a 2-D line pattern image;

a database configured to store at least one previously obtained 2-D facial fingerprint; and

a controller communicatively connected to said imager, wherein said controller is configured to compare said 2-D facial fingerprint to said at least one previously obtained 2-D facial fingerprint stored in said database, such that said controller classifies the occupant as one of a forward facing classification, a rightward facing classification, a leftward facing classification, an upward classification, a downward classification, an attentive classification, and a non-attentive classification.

12. The system of claim 11, wherein said the light source projects an initial line pattern at the target, such that the target is at least a portion of a face of a user of the image classification system.

13. The system of claim 11, wherein said at least one light source is a plurality of light sources, and at least a portion of said plurality of light sources is configured to emit light at different wavelengths.

14. The system of claim 11 further comprising a 2-D line pattern generator in communicative connection with said imager, wherein said imager is configured to obtain an image of the target, and said 2-D line pattern generator is configured to produce said obtained 2-D line pattern as a function of said obtained image.

15. The system of claim 14, wherein said 2-D line pattern is less data bytes than said image obtained by said imager of the target.

16. A method of classifying the target in an image, said method comprising the steps of:

illuminating the target;

obtaining an image based upon at least a portion of light reflected by said illumination of the target, wherein said image is an obtained two-dimensional (2-D) line pattern that is representative of at least a portion of said light reflected by the target;

comparing said 2-D line pattern to at least one previously obtained 2-D line pattern; and

classifying the target in said image based upon said comparison of said 2-D line pattern and said at least one previously obtained 2-D line pattern.

17. The method of claim 16, wherein said 2-D line pattern is a facial fingerprint.

18. The method of claim 17, wherein said 2-D facial fingerprint is a function of at least a portion of a face of an occupant in a vehicle, such that said 2-D line pattern has a parameter of a 2-D line pattern image.

19. The method of claim 16, wherein said step of classifying the target further comprises classifying the target as one of forward facing, rightward facing, leftward facing, upward facing, downward facing, attentive, and non-attentive.

20. The method of claim 16, wherein said 2-D line pattern is less data bytes than said image obtained of the target.