Object detection apparatus object detection method and storage medium

1. An object detection apparatus comprising:

a circuit which serves as:

a first setting unit configured to set a first partial region in an input image;

a second setting unit configured to set a second partial region in the input image, the second partial region being outside the first partial region set by the first setting unit;

a third setting unit configured to set a third partial region, the third partial region being inside the second partial region and containing the first partial region;

a salience degree deriving unit configured to derive a salience degree based on a feature amount in the first partial region and a feature amount in the second partial region; and

a detection unit configured to detect a main object in the input image based on the salience degree derived by the salience degree deriving unit and an information amount in the third partial region set by the third setting unit.

2. The object detection apparatus according to claim 1 , further comprising:

the circuit serving as a score deriving unit configured to derive a score according to a salience degree derived by the salience degree deriving unit and the information amount in the third partial region set by the third setting unit,

wherein the detection unit detects the main object in the input image based on the score derived by the score deriving unit.

3. The object detection apparatus according to claim 1 , wherein the second setting unit sets the second partial region that contains the first partial region.

4. The object detection apparatus according to claim 3 , wherein the information amount in the second partial region is entropy of the feature amount in the second partial region.

5. The object detection apparatus according to claim 3 , wherein the information amount in the second partial region is an output value of the feature amount in the second partial region from a bandpass filter.

6. The object detection apparatus according to claim 3 , wherein the salience degree deriving unit derives the salience degree based on a difference in statistical feature amount distribution between the first partial region and the second partial region.

7. The object detection apparatus according to claim 3 , wherein the detection unit generates a score map in which scores derived from a plurality of partial regions in the input image by the score deriving unit are arranged, and detects the main object in the input image based on a threshold value for the score map.

8. The object detection apparatus according to claim 7 , wherein the threshold value is determined through learning.

9. The object detection apparatus according to claim 1 , wherein the information amount in the third partial region is a gradient intensity of the feature amount in the third partial region.

10. A method for detecting an object, the method comprising:

setting a first partial region in an input image;

setting a second partial region in the input image, the second partial region being outside the first partial region set in the setting of the first partial region;

setting a third partial region, the third partial region being inside the second partial region and containing the first partial region;

deriving a salience degree based on a feature amount in the first partial region and a feature amount in the second partial region; and

detecting a main object in the input image based on the salience degree derived in the deriving and an information amount in the third partial region set in the setting of the third partial region.

11. A non-transitory computer-readable storage medium storing a program that causes a computer to execute a method for detecting an object, the method comprising:

setting a first partial region in an input image;

setting a second partial region in the input image, the second partial region being outside the first partial region set in the setting of the first partial region;

setting a third partial region, the third partial region being inside the second partial region and containing the first partial region;

deriving a salience degree based on a feature amount in the first partial region and a feature amount in the second partial region; and

detecting a main object in the input image based on the salience degree derived in the deriving and an information amount in the third partial region set in the setting of the third partial region.

12. The object detection apparatus according to claim 1 , wherein the third setting unit sets the third partial region based on an information amount in the second partial region.

13. An object detection apparatus comprising:

a memory; and

a processor in communication with the memory, the processor configured to provide:

a first setting unit configured to set a first partial region in an input image;

a second setting unit configured to set a second partial region in the input image, the second partial region being outside the first partial region set by the first setting unit;

a third setting unit configured to set a third partial region, the third partial region being inside the second partial region and containing the first partial region;

a salience degree deriving unit configured to derive a salience degree based on a feature amount in the first partial region and a feature amount in the second partial region; and

a detection unit configured to detect a main object in the input image based on the salience degree derived by the salience degree deriving unit and an information amount in the third partial region set by the third setting unit.