Device and method for service provision according to prepared reference images to detect target object
A non-transitory recording medium storing a program that causes a computer to execute a process, the process includes: imaging a given object from plural different angles, and extracting from the plural obtained captured images, one or plural captured images having a feature amount that differs from a feature amount in another captured image by more than a specific reference amount; and providing the one or the plural extracted captured images as determination-use images employable in determination as to whether or not the given object is included in a captured image.
1. A non-transitory recording medium storing a program that causes a computer to execute a process, the process comprising:
extracting a plurality of reference images from a plurality of first captured images obtained by imaging a given object from a plurality of different angles,
at least one extracted reference image, among the extracted reference images, has a feature amount that differs by more than a reference amount from feature amounts of other extracted reference image; and
providing the extracted reference images as determination—use images employable to determine whether the given object is included in a second captured image.
2. A non-transitory recording medium storing a program that causes a computer to execute a process, the process comprising:
from a plurality of first captured images obtained by imaging a given object from a plurality of different angles, extracting and storing, in a memory, a plurality of reference images, at least one extracted reference image, among the extracted reference images, has a feature amount that differs by more than a reference amount from feature amounts of other extracted reference image; and
comparing a feature amount of a second captured image other than the first captured images against the feature amount of the extracted reference images stored in the memory; and
determining whether the given object is included in the second captured image based on a comparison result of the comparing.
3. The non-transitory recording medium of claim 2 , wherein the second captured image is acquired through the Internet.
4. The non-transitory recording medium of claim 2 , wherein the process further comprises:
identifying a provision source of the second captured image in which the given object has been determined to be included; and
outputting a message to the identified provision source.
5. The non-transitory recording medium of claim 4 , wherein the process further comprises:
outputting information that identifies the provision source of the second captured image in which the given object has been determined to be included.
6. The non-transitory recording medium of claim 1 , wherein the extracting further comprises:
setting a first captured image, among the first captured images, as a reference image;
extracting another first captured image, among the first captured images, as the at least one extracted reference image which has a feature amount that differs from a feature amount of the set reference image by more than the reference amount; and
setting the at least one extracted reference image as a new reference image, and further extracting another first captured image, among the first captured images, as another extracted reference image which has a feature amount that differs from a feature amount of the set new reference image by more than the reference amount.
7. The non-transitory recording medium of claim 1 , wherein the process further comprises:
categorizing the first captured images into a plurality of classes such that typical vectors of the classes differ by more than the reference amount from each other; and
extracting the first captured images from the classes.
8. The non-transitory recording medium of claim 1 , wherein the first captured images are first captured images in which a background behind the given object has been eliminated.
9. A service provision method, comprising:
by at least one processor,
extracting a plurality of reference images from a plurality of first captured images obtained by imaging a given object from a plurality of different angles,
at least one extracted reference image, among the extracted reference images, has a feature amount that differs by more than a reference amount from feature amounts of other extracted reference image; and
providing the extracted reference images as determination—use images employable to determine whether the given object is included in a second captured image.
10. The service provision method of claim 9 , wherein the extracting further comprises:
setting a first captured image from among the first captured images as a reference image
extracting another first captured image, among the first captured images, as the at least one extracted reference image which has a feature amount that differs from a feature amount of the set reference image by more than the reference amount; and
setting the extracted at least one reference image as a new reference image, and further extracting another first captured image, among the first captured images, as another extracted reference image which has a feature amount that differs from a feature amount of the set new reference image by more than the reference amount.
11. The service provision method of claim 9 , further comprising:
by the at least one processor,
categorizing the first captured images into a plurality of classes such that typical vectors of the classes differ by more than the reference amount from each other; and
extracting the first captured images from the classes.
12. The service provision method of claim 9 , wherein the first captured images are captured images in which a background behind the given object has been eliminated.