INFORMATION PROCESSOR, INFORMATION PROCESSING METHOD, AND PROGRAM

• US 20150324971A1
• Filed: 07/21/2015
• Published: 11/12/2015
• Est. Priority Date: 09/05/2011
• Status: Active Grant

• Current Assignee: C/O SONY CORPORATION
• Original Assignee: C/O SONY CORPORATION
• Inventors: Sato, Tatsuhito, Gotoh, Tomohiko, Mochizuki, Daisuke, Mochizuki, Shunsuke, Sun, Yun, Okamura, Yuki, Yaeda, Takeshi
• Granted Patent: US 9,589,341 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 18/2411 based on the proximity to a decision surface, e.g. support vector machines
• CPC Class Codes: G06T 11/60 Editing figures and text; Combining figures or text
• CPC Class Codes: G06T 2207/10004 Still image; Photographic image
• CPC Class Codes: G06T 2207/30128 Food products
• CPC Class Codes: G06T 7/0004 Industrial image inspection
• CPC Class Codes: G06V 10/50 by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis

Abstract

An information processor includes a detection unit detecting a photographic subject region of an image, a characteristic amount generation unit generating a characteristic amount including at least positional information of the photographic subject region for each of the detected photographic subject region, a combined characteristic amount generation unit generating a combined characteristic amount corresponding to the image by combining the characteristic amount generated for each of the photographic subject region, and an identification unit identifying a label corresponding to a combination of a photographic subject appearing in the image based on the generated combined characteristic amount.

Claims

1. An information processor comprising:

circuitry configured to;

detect an object in a plurality of individual regions of an image;

generate a respective characteristic amount identifying a subject supported by each of the individual regions detected;

generate a combined characteristic amount corresponding to the image, wherein the combined characteristic amount is generated by combining characteristic amounts for at least two of the subjects supported by individual regions having the largest relative size; and

determine a label corresponding to a combination of subjects appearing in the image based on the generated combined characteristic amount.

2. The information processor according to claim 1, wherein the circuitry is further configured to carry out machine learning using the label and the combined characteristic amount corresponding to an image for learning as inputs.

3. The information processor according to claim 1, wherein the circuitry is further configured to generate the characteristic amount by combining at least positional information of the individual regions and an identification result of the each subject for each of the detected individual regions.

4. The information processor according to claim 1, wherein the circuitry is further configured to generate the characteristic amount by combining at least positional information of the individual regions and an image characteristic of the subject supported by each individual detected region.

5. The information processor according to claim 1, wherein the circuitry is further configured to detect a shape of the plurality of individual regions and positional information is detected as being center coordinates for each of the individual regions.

6. The information processor according to claim 1, wherein

the image is a cooked food image having a cooked food as the subject supported by each of the detected individual regions, and

the circuitry is further configured to detect individual plates in the cooked food image as the individual regions.

7. An image processing method comprising:

detecting at least a relative size and positional information for a plurality of individual regions of an image, each of the individual regions appearing in the image as individually supporting one of a plurality of subjects;

generating a characteristic amount identifying the subject supported by each of the individual regions detected in the detecting step;

generating a combined characteristic amount corresponding to the image, wherein the combined characteristic amount is generated by combining the characteristic amount generated for at least two subjects supported by the individual regions having a largest relative size; and

determining a label corresponding to a combination of a subjects appearing in the image based on the generated combined characteristic amount.

8. The method according to claim 7, wherein generating the characteristic amount includes combining at least the positional information of the individual regions and an identification result of each individual subject for each of the detected individual regions.

9. The method according to claim 7, wherein generating the characteristic amount includes combining at least the positional information of the individual regions and an image characteristic of the subject-supported by each individual detected region.

10. The method according to claim 7, wherein the detecting step further detects a shape for the plurality of individual regions and the positional information is detected as being center coordinates for each of the individual regions.

11. The method according to claim 7, wherein the image is a cooked food image having a cooked food as each subject supported by each of the detected individual regions, and the detecting step detects individual plates in the cooked food image as the individual regions.

12. A non-transitory computer-readable medium having a set of computer-executable instructions embodied thereon that, when executed, cause a computer to perform a method comprising:

detecting at least a relative size and positional information for a plurality of individual regions of an image, each of the individual regions appearing in the image as individually supporting one of a plurality of subjects;

generating a characteristic amount identifying the subject supported by each of the individual regions;

generating a combined characteristic amount corresponding to the image, wherein the combined characteristic amount is generated by combining the characteristic amount generated for at least two subjects supported by the individual regions having a largest relative size; and

determining a label corresponding to a combination of a subjects appearing in the image based on the generated combined characteristic amount.