US 12,086,952 B2
Image processing apparatus, conversion apparatus, image processing method, conversion method and program
Satoshi Suzuki, Musashino (JP); Motohiro Takagi, Musashino (JP); Kazuya Hayase, Musashino (JP); Takayuki Onishi, Musashino (JP); and Atsushi Shimizu, Musashino (JP)
Assigned to NIPPON TELEGRAPH AND TELEPHONE CORPORATION, Tokyo (JP)
Appl. No. 17/596,139
Filed by NIPPON TELEGRAPH AND TELEPHONE CORPORATION, Tokyo (JP)
PCT Filed Jun. 6, 2019, PCT No. PCT/JP2019/022644
§ 371(c)(1), (2) Date Dec. 3, 2021,
PCT Pub. No. WO2020/246012, PCT Pub. Date Dec. 10, 2020.
Prior Publication US 2022/0237734 A1, Jul. 28, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 3/40 (2024.01); G06V 20/50 (2022.01)
CPC G06T 3/40 (2013.01) [G06V 20/50 (2022.01); G06T 2207/20081 (2013.01)] 19 Claims
OG exemplary drawing
 
1. An image processing device, comprising:
an image processing unit configured to execute image processing on an image based on an input image, and output a result of the image processing, wherein
the input image is a post-conversion image obtained by performing image conversion on an original image, and
the image conversion includes image conversion for further reducing a data size of the original image while maintaining a feature acquired from the original image and related to an object similar to a subject captured in the original image and maintaining processing accuracy of the image processing,
wherein the image conversion includes converting the input image so as to satisfy each of
a first condition that the input image is converted to an image having a data size smaller than that of the input image,
a second condition that, when a subject captured in the input image is an object belonging to a first type, and there is each degree of certainty that the type to which the subject belongs is recognized as a first type, a second type different from the first type, and a third type different from the first type and the second type,
wherein the input image is converted to an image in which a feature indicating at least the second type is maintained,
and the image processing unit is further configured to:
recognize a subject captured in post-conversion image, and
reduce weighting for the second condition relative to weighting for the first condition when there is a degree of certainty that the subject is recognized as the first type.