Object determining device and imaging apparatus

• US 7,912,363 B2
• Filed: 03/16/2009
• Issued: 03/22/2011
• Est. Priority Date: 07/15/2003
• Status: Active Grant

• Current Assignee: Omron Corporation
• Original Assignee: Omron Corporation
• Inventors: Imae, Fumikazu, Yamamoto, Masashi, Minato, Yoshihisa, Tsuyuguchi, Takuya
• Primary Examiner(s): FULLER, RODNEY EVAN
• Application Number:
  US12/404,908
• Publication Number:
  US 20090180696A1
• Time in Patent Office: 736 Days
• Field of Search: 396/123
• US Class Current: 396/123
• CPC Class Codes: G03B 13/36 Autofocus systems
• CPC Class Codes: G03B 7/091 Digital circuits
• CPC Class Codes: G06V 40/165 using facial parts and geometric relationships
• CPC Class Codes: H04N 23/611 where the recognised objects include parts of the human body
• CPC Class Codes: H04N 23/70 Circuitry for compensating brightness variation in the scene

Abstract

Face detection is executed only on an image imaged based on an arbitrary (predetermined) focal point information defined in advance. In face detection, the face of a person is detected based on a relative value of statistics in a plurality of characteristic regions produced by contour or parts of the face of the person. Thus, even if the face in the image used for face detection is blurred, the statistics in a certain region can be acquired, and thus such face is detectable. Therefore, the primary subject to be focused can be rapidly detected without performing focal point control at a stage of preliminary imaging in the imaging apparatus.

Claims

1. An object determining device comprising:

a detecting part for detecting a face of a person from an input image;

a classifying part for classifying, when a plurality of faces of people are detected by the detecting part, the plurality of detected faces of people into a plurality of clusters;

a temporary determining part for determining a face to be focused and/or face on which to perform exposure control when performing imaging from faces of people in the cluster for each of the plurality of clusters; and

an ultimate determining part for ultimately determining the target face from the faces determined by the temporary determining part.

2. The object determining device of claim 1, wherein the detecting part detects a face of a person for each of a plurality of blocks divided from the input image;

wherein the object determining device further comprises a block determining part for determining a selected block for determining a face to be focused and/or face on which to perform exposure control when performing imaging based on a detection result of the detecting part; and

wherein the classifying part classifies the plurality of detected faces into a plurality of clusters for the selected block determined by the block determining part.

3. The object determining device of claim 1, wherein the temporary determining part further comprises:

a deciding part for deciding a largest face from the detected faces of people in the cluster for each of the plurality of clusters; and

a selecting part for selecting the largest face and a face having a size within a predetermined range with the size of the largest face as a reference from the detected faces in the cluster;

wherein the temporary determining part determines the face to be focused and/or face on which to perform exposure control when performing imaging from the faces selected by the selecting part.

4. An object determining device comprising:

a frame acquiring section for acquiring an image of a predetermined frame as an image to be processed from an input time-series image including a plurality of frames;

a detecting part for detecting a face of a person from the image to be processed;

a classifying part for classifying, when a plurality of faces of people are detected by the detecting part, the plurality of detected faces of people into a plurality of clusters;

a temporary determining part for determining a face to be focused and/or face on which to perform exposure control when performing imaging from faces of people included in the cluster for each of the plurality of clusters; and

an ultimate determining part for ultimately determining the target face from faces determined by the temporary determining part.

5. The object determining device of claim 4, wherein the detecting part detects a face of a person for each of a plurality of blocks divided from the image to be processed;

wherein the object determining device further comprises a block determining part for determining a selected block for determining a face to be focused and/or face on which to perform exposure control when performing imaging based on a detection result of the detecting part; and

wherein the classifying part classifies the plurality of detected faces into a plurality of clusters for the selected block determined by the block determining part.

6. The object determining device of claim 4, wherein the temporary determining part further comprises:

a deciding part for deciding a largest face from the detected faces of people in the cluster for each of the plurality of clusters; and

a selecting part for selecting the largest face and a face having a size within a predetermined range with the size of the largest face as a reference from the detected faces in the cluster;

wherein the temporary determining part determines the face to be focused and/or face on which to perform exposure control when performing imaging from the faces selected by the selecting part.

7. An object determining device comprising:

a detecting part for detecting a face of a person from an input image;

a classifying part for classifying, when a plurality of faces of people are detected by the detecting part, the plurality of detected faces of people into a plurality of clusters based on respective positions; and

a cluster determining part for determining a selected cluster to determine a face to be focused and/or face on which to perform exposure control when performing imaging from the plurality of clusters; and

a determining part for determining the target face from faces included in the selected cluster determined by the cluster determining part.

8. An object determining device comprising:

a frame acquiring section for acquiring an image of a predetermined frame as an image to be processed from an input time-series image including a plurality of frames;

a detecting part for detecting a face of a person from the image to be processed;

a classifying part for classifying, when a plurality of faces of people are detected by the detecting part, the plurality of detected faces of people into a plurality of clusters based on the respective positions;

a cluster determining part for determining a selected cluster to determine a face to be focused and/or face on which to perform exposure control when performing imaging from the plurality of clusters; and

a determining part for determining the target face from faces included in the selected cluster determined by the cluster determining part.

9. A non-transitory computer readable medium containing a program for an information processing device, said program when executed by the processing device causes the processing device to perform a method comprising the steps of:

detecting a face of a person from an input image;

classifying, when a plurality of faces of people are detected, the plurality of detected faces of people into a plurality of clusters based on the respective positions;

determining a face to be focused and/or face on which to perform exposure control when performing imaging from faces included in the cluster for each of the plurality of clusters; and

ultimately determining the target face from faces detected in each of the cluster.

10. The non-transitory computer readable medium of claim 9, wherein the detecting step occurs in each of a plurality of blocks divided from the input image;

wherein the program further causes the processing device to perform the step of determining a selected block to determine a face to be focused and/or face on which to perform exposure control when performing imaging based on detection result in each of the block; and

wherein the classifying step classifies the plurality of detected faces into a plurality of clusters for the selected block determined by the determining step.

11. A non-transitory computer readable medium containing a program for an information processing device, said program when executed by the processing device causes the processing device to perform a method comprising the steps of:

detecting a face of a person from an input image;

classifying, when a plurality of faces of people are detected, the plurality of detected faces of people into a plurality of clusters based on the respective positions;

determining a selected cluster for determining a face to be focused and/or face on which to perform exposure control when performing imaging from the plurality of clusters; and

determining the target face from faces included in the determined selected cluster.

12. The non-transitory computer readable medium of claim 11, said method further comprising the steps of:

deciding which of the plurality of faces is a largest face out of the plurality of detected faces of people in the selected cluster; and

selecting the largest face and at least one other face having a size within a predetermined range with the size of the largest face as a reference from the detected faces;

wherein the determining step determines the target face to be focused and/or face on which to perform exposure control when performing imaging from the faces selected by the selecting part.

13. An object determining device comprising:

a detecting part for detecting a face of a person from an input image; and

a determining part for determining, when a plurality of faces of people are detected, one of the plurality of faces as a subject to be focused and/or subject on which to perform exposure control when performing imaging,

wherein the determining part determines the subject to be focused and/or subject on which to perform exposure control when performing imaging by using one selected from the group consisting of;

classifying the plurality of detected faces into a plurality of clusters, from which one of the plurality of clusters is selected, and selecting the subject to be focused and/or subject on which to perform exposure control when performing imaging as one of the detected faces from the selected cluster;

dividing the input image into a plurality of blocks, from which one of the plurality of blocks is selected, and selecting the subject to be focused and/or subject on which to perform exposure control when performing imaging as one of the detected faces from the selected block; and

selecting the largest face from the plurality of detected faces and at least one other face having a size within a predetermined range with the size of the largest face as a reference from the detected faces, and selecting the subject to be focused and/or subject on which to perform exposure control when performing imaging as one of the selected faces.