US 12,437,537 B2
Image processing apparatus, image processing method, and storage medium that estimate concepts to select a type of feature amount based on the estimated concepts and determine a weight of the selected type
Takayuki Yamada, Kanagawa (JP); Kazuya Ogasawara, Kanagawa (JP); and Shinjiro Hori, Kanagawa (JP)
Assigned to CANON KABUSHIKI KAISHA, Tokyo (JP)
Filed by CANON KABUSHIKI KAISHA, Tokyo (JP)
Filed on Nov. 23, 2021, as Appl. No. 17/533,864.
Claims priority of application No. 2020-198582 (JP), filed on Nov. 30, 2020.
Prior Publication US 2022/0172472 A1, Jun. 2, 2022
Int. Cl. G06V 20/30 (2022.01); G06T 7/00 (2017.01); G06T 11/60 (2006.01); G06V 40/16 (2022.01)
CPC G06V 20/30 (2022.01) [G06T 7/0002 (2013.01); G06T 11/60 (2013.01); G06T 2207/30168 (2013.01); G06V 40/161 (2022.01)] 17 Claims
OG exemplary drawing
 
1. An image processing apparatus comprising:
at least one processor coupled to a memory, serving as:
a first designation unit configured to designate a first image group;
a second designation unit configured to designate a second image group;
an analysis unit configured to analyze each image included in the first image group and the second image group and to output feature amounts;
a first selection unit configured to select a type of feature amount to be used, based on the second image group, from among the feature amounts that are output by the analysis unit, wherein the first selection unit estimates a plurality of concepts to select the type of feature amount based on the estimated plurality of concepts;
a determination unit configured to determine a score reference that gives a score to an image, based on a feature amount of the selected type out of feature amounts of the second image group, wherein the determination unit determines a weight of a feature amount of the type selected by the first selection unit based on the plurality of concepts, and changes the score reference based on the weight of the feature amount of the selected type to produce a changed score reference;
a scoring unit configured to perform scoring of each image of the first image group, based on feature amounts of the first image group and the changed score reference, to produce a first image group score, which is a score corresponding to each image of the first image groups; and
a second selection unit configured to select an image from the first image group, based on the first image group score obtained by the scoring,
wherein the determination unit determines a first score reference in a case when a standard deviation of feature amounts of the second image group is low, and determines a second score reference in a case when the standard deviation is great,
wherein, based on the first score reference, a first feature amount is scored higher than a second feature amount that is far from an average value of the feature amounts of the second image group than is the first feature amount, and
wherein, based on the second score reference, the first feature amount and the second feature amount are scored such that a difference between a score of the first feature amount and a score of the second feature amount is less than that determined based on the first score reference.