Information processor and information processing method

1. An information processor comprising:

an image device arranged to capture a shot image and to transform the shot image into a plurality of depth images having different resolutions;

a candidate area extraction section arranged to extract candidate areas in which a target picture is likely to exist in the shot image by performing template matching on each of the plurality of depth images using a template image, the depth images representing, on an image plane, a distance of a subject from a camera in a depth direction as a pixel value, the template image representing a shape and a size of the target to be analyzed,

wherein, for each of the plurality of depth images, the template image is varied in size during template matching only within a predefined magnification range;

a detailed information acquisition section arranged to analyze the candidate areas extracted by the candidate area extraction section; and

an output information generation section arranged to generate output information based on results of image analysis performed by the detailed information acquisition section and output the output information.

2. The information processor of claim 1 , wherein

the detailed information acquisition section determines, as a target picture area and in each frame of a movie, i.e., the shot image, an area of the associated depth image, the area inside the candidate area extracted by the candidate area extraction section, the area being made up of pixels whose values fall within a given range, so as to estimate the target picture area at the next frame for tracking.

3. The information processor of claim 1 , wherein

the detailed information acquisition section performs tracking of the target in the movie, i.e., the shot image, and corrects tracking results obtained for each frame of the movie based on the candidate area for the target extracted by the candidate area extraction section from the associated depth image.

4. The information processor of claim 1 , wherein

the candidate area extraction section extracts, from each of the plurality of depth images, the candidate area for the picture existing in a distance range in the depth direction associated with one of the resolutions, and

the detailed information acquisition section performs image analysis on the shot image with the resolution associated with that of the depth image from which the candidate area has been extracted.

5. The information processor of claim 1 , wherein

the image analysis is facial recognition analysis, and

the resolution of the depth image from which the candidate area is extracted by the candidate area extraction section is smaller than that of the shot image on which image analysis is performed by the detailed information acquisition section using information relating to the candidate area.

6. The information processor of claim 1 , wherein

the detailed information acquisition section requests the camera to adjust an exposure value in such a manner that if a luminance of the candidate area extracted by the candidate area extraction section in the shot image falls outside a proper range, the luminance falls within the proper range.

7. The information processor of claim 1 , wherein

the candidate area extraction section extracts the candidate area for human head, i.e., the target, and

the detailed information acquisition section detects a facial area by performing a face detection process on the extracted candidate area in the shot image and identifies the person of the picture by performing a face recognition process on the detected facial area using a shot image with a higher resolution than that of the shot image used for the face detection process.

8. An information processing method used by an information processor to process information using a shot image from a camera, the information processing method comprising:

transforming the shot image into a plurality of depth images having differing resolutions,

extracting candidate areas in which a target picture is likely to exist in the shot image by performing template matching on each of the plurality of depth images using a template image, the depth images representing, on an image plane, a distance of a subject from the camera in a depth direction as a pixel value, the template image being read from a memory and representing a shape and a size of the target to be analyzed;

wherein, for each of the plurality of depth images, the template image is varied in size during template matching only within a predefined magnification range;

analyzing the extracted candidate areas; and

generating an output image reflecting the image analysis results and outputting the output image to a display device.

9. A computer program for a computer stored on a non-transitory computer readable medium for causing a computer to execute:

transforming a shot image into a plurality of depth images having differing resolutions;

extracting a candidate area in which a target picture is likely to exist in the shot image by performing template matching on each of the plurality of depth images using a template image, the depth images representing, on an image plane, a distance of a subject from a camera in a depth direction as a pixel value, the template image being read from a memory and representing a shape and a size of the target to be analyzed;

wherein, for each of the plurality of depth images, the template image is varied in size during template matching only within a predefined magnification range;

analyzing extracted candidate areas; and

generating an output image reflecting the image analysis results and outputting the output image to a display device.