Image processing device, image processing method, and image processing program

1. An image processing device comprising:

a memory storing a software component; and

at least one processor configured to execute the software component to perform:

obtaining features in each of scaled samples of a region of interest in a probe image;

computing probabilities of pixels in the scaled samples that contribute to a score or a label of an object of interest in the region;

removing features from the scaled samples which are not essential for computing the score or the label of the object, using the computed probabilities; and

selecting the scaled sample using remaining features after removing the features.

2. The image processing device according to claim 1 ,

wherein the at least one processor further configured to execute the software component to perform:

obtaining a similarity between a given target image and a scaled sample of the probe image and selecting the scaled sample with a maximum similarity as a final output.

3. The image processing device according to claim 2 ,

wherein the at least one processor further configured to execute the software component to perform:

learning a models parameters by one or more series of training samples which contain target image and probe image pairs and a label indicating whether they are the same object or not.

4. The image processing device according to claim 3 ,

wherein the at least one processor further configured to execute the software component to perform:

normalizing the remaining features again after removing the features.

5. The image processing device according to claim 2 ,

wherein the at least one processor further configured to execute the software component to perform:

generating a mask for removing the features which are not essential for the computing the score or the label of the object, using the computed probabilities, and removes the features from the scaled samples, using the generated mask.

6. The image processing device according to claim 5 ,

wherein the at least one processor further configured to execute the software component to perform:

updating feature map by applying the mask generated for removing the features whose pixels result in a saliency map with low probability.

7. The image processing device according to claim 6 ,

wherein the at least one processor further configured to execute the software component to perform:

normalizing the remaining features again after removing the features.

8. The image processing device according to claim 5 ,

wherein the at least one processor further configured to execute the software component to perform:

normalizing the remaining features again after removing the features.

9. The image processing device according to claim 2 ,

wherein the at least one processor further configured to execute the software component to perform:

normalizing the remaining features again after removing the features.

10. The image processing device according to claim 1 ,

wherein the at least one processor further configured to execute the software component to perform:

generating a mask for removing the features which are not essential for the computing the score or the label of the object, using the computed probabilities, and removing the features from the scaled samples, using the generated mask.

11. The image processing device according to claim 10 ,

wherein the at least one processor further configured to execute the software component to perform:

updating feature map by applying the mask generated for removing the features whose pixels result in a saliency map with low probability.

12. The image processing device according to claim 11 ,

wherein the at least one processor further configured to execute the software component to perform:

normalizing the remaining features again after removing the features.

13. The image processing device according to claim 10 ,

wherein the at least one processor further configured to execute the software component to perform:

normalizing the remaining features again after removing the features.

14. The image processing device according to claim 1 ,

wherein the at least one processor further configured to execute the software component to perform:

normalizing the remaining features again after removing the features.

15. An image processing method comprising:

obtaining features in each of scaled samples of a region of interest in a probe image;

computing probabilities of pixels in the scaled samples that contribute to a score or a label of an object of interest in the region;

removing features from the scaled samples which are not essential for computing the score or the label of the object, using the computed probabilities; and

selecting the scaled sample using remaining features after removing the features.

16. The image processing method according to claim 15 , further comprising:

obtaining a similarity between a given target image and a scaled sample of the probe image; and

selecting the scaled sample with a maximum similarity as a final output.

17. A non-transitory computer-readable recording medium having recorded therein an image processing program that, when executed by a computer,

obtains features in each of scaled samples of a region of interest in a probe image,

computes probabilities of pixels in the scaled samples that contribute to a score or a label of an object of interest in the region,

removes features from the scaled samples which are not essential for the computing the score or the label of the object, using the computed probabilities, and

selects the scaled sample using remaining features after removing the features.

18. A non-transitory computer-readable recording medium according to claim 17 , the image processing program when executed by the computer,

obtains a similarity between a given target image and a scaled sample of the probe image, and

selects the scaled sample with a maximum similarity as a final output.