Systems and methods for face detection and recognition using infrared imaging

• US 20050226471A1
• Filed: 03/28/2005
• Published: 10/13/2005
• Est. Priority Date: 03/29/2004
• Status: Active Grant

• Current Assignee: Continental Automotive GmbH (Continental AG), Siemens Healthineers AG (Siemens AG)
• Original Assignee: Siemens Corp. (Siemens AG)
• Inventors: Okada, Kazunori, Comaniciu, Dorin, Bascle, Benedicte, Baratoff, Gregory, Kohler, Thorsten, Singh, Maneesh
• Granted Patent: US 7,542,592 B2
• Time in Patent Office: Days
• US Class Current: 382/118
• CPC Class Codes: G06V 40/161 Detection; Localisation; Normalisation
• CPC Class Codes: G06V 40/172 Classification, e.g. identification

Abstract

Methods for image processing for detecting and recognizing an image object include detecting an image object using pose-specific object detectors, and performing fusion of the outputs from the pose-specific object detectors. The image object is recognized using pose-specific object recognizers that use outputs from the pose-specific object detectors and the fused output of the pose-specific object detectors; and by performing fusion of the outputs of the pose-specific object recognizers to recognize the image object.

Claims

1. A method of image processing for detecting and recognizing an image object in an image, the method comprising:

detecting at least one image object using a plurality of pose-specific object detectors;

fusing the outputs from the object pose detectors;

recognizing at least one image object using a plurality of pose-specific object recognizers that use outputs from the pose-specific detectors and the fused output of the pose-specific detectors; and

fusing the outputs of the pose-specific object recognizers to recognize the image object.

2. The method of claim 1, wherein the image is an infrared image obtained using a camera that can capture an infrared image.

3. The method of claim 1, wherein each one of the pose-specific object detector detects and fuses output of a plurality of component detectors which detect components of a given pose of an image object.

4. The method of claim 3, wherein the image object is a face and the component detector detects at least one of a left-eye component, a right-eye component and a lower-face face component.

5. The method of claim 4, wherein the step of performing fusion of the pose-specific object detector output comprises:

selecting the output of the pose-specific object detector with a highest confidence measure among the pose-specific object detectors.

6. The method of claim 5, wherein the step of performing fusion of the pose-specific object recognizer outputs comprises:

recognizing the face image with one of the pose-specific object recognizers that uses the output of the pose-specific object detector with the highest confidence measure.

7. The method of claim 4, wherein the component detectors comprise a front view detector and a side view detector, and the image recognizers comprise a front view recognizer and a side view recognizer.

8. The method of claim 4, wherein the pose-specific object detectors form an array of pose-specific object detectors for detecting faces having a plurality of poses in the input image.

9. The method of claim 8, wherein the pose-specific object recognizers form an array of pose-specific object recognizers for recognizing a plurality of poses of the face.

10. A method for detecting and recognizing a face in an image, the method comprising:

detecting one or more faces in the image using a plurality of pose-specific detectors, wherein the face poses are sourced from an infrared image of the face;

fusing the outputs from the pose-specific detectors;

recognizing the face using a plurality of pose-specific recognizers that use outputs from the pose-specific detectors and the fused output of the pose-specific detectors;

fusing the outputs of the pose-specific recognizers; and

recognizing the face from an image database of multiple faces using the fused output of the pose-specific recognizers.

11. The method of claim 10, wherein each one of the pose detector detects and fuses output of a plurality of component detectors that detect components of a given face pose.

12. The method of claim 11, wherein the components of a given face pose comprise a left-eye component, a right-eye component and a lower-face face component.

13. A program storage device readable by machine, tangibly embodying a program of instructions executable by the machine to perform method steps comprising:

detecting at least one image object using a plurality of pose-specific object detectors;

fusing the outputs from the object pose detectors;

recognizing at least one image object using a plurality of pose-specific object recognizers that use outputs from the pose-specific detectors and the fused output of the pose-specific detectors; and

fusing the outputs of the pose-specific object recognizers to recognize the image object.

14. The device of claim 13, wherein the image is an infrared image obtained using a camera that can capture an infrared image.

15. The device of claim 13, wherein each one of the pose-specific object detector detects and fuses output of a plurality of component detectors which detect components of a given pose of an image object.

16. The device of claim 15, wherein the image object is a face and the component detector detects at least one of a left-eye component, a right-eye component and a lower-face face component.

17. The device of claim 16, wherein instructions for the step of performing fusion of the pose-specific object detector output comprise:

selecting the output of the pose-specific object detector with a highest confidence measure among the pose-specific object detector.

18. The device of claim 17, wherein instructions for the step of performing fusion of the pose-specific object recognizer outputs comprise:

recognizing the face image with one of the pose-specific object recognizers that uses the output of the pose-specific object detector with the highest confidence measure.

19. The device of claim 16, wherein the component detectors comprise a front view detector and a side view detector, and the image recognizers comprise a front view recognizer and a side view recognizer.

20. The device of claim 16, wherein the pose-specific object detectors form an array of pose-specific object detectors for detecting a plurality of poses of the face.

21. The method of claim 20, wherein the pose-specific object recognizers form an array of pose-specific object recognizers for recognizing a plurality of poses of the face.