US 12,175,720 B2
Feature quantity extracting device, feature quantity extracting method, identification device, identification method, and program
Tasuku Sano, Azumino (JP); Ryuhei Okuyama, Matsumoto (JP); Akane Minagawa, Matsumoto (JP); Yoshihiro Teshima, Higashimurayama (JP); and Akira Hamada, Sagamihara (JP)
Assigned to SHINSHU UNIVERSITY, Nagano (JP); and CASIO COMPUTER CO., LTD., Tokyo (JP)
Appl. No. 17/620,672
Filed by SHINSHU UNIVERSITY, Matsumoto (JP); and CASIO COMPUTER CO., LTD., Tokyo (JP)
PCT Filed Mar. 25, 2020, PCT No. PCT/JP2020/013225
§ 371(c)(1), (2) Date Dec. 17, 2021,
PCT Pub. No. WO2020/255517, PCT Pub. Date Dec. 24, 2020.
Claims priority of application No. 2019-113990 (JP), filed on Jun. 19, 2019; and application No. 2020-043778 (JP), filed on Mar. 13, 2020.
Prior Publication US 2022/0245919 A1, Aug. 4, 2022
Int. Cl. G06V 10/20 (2022.01); A61B 5/00 (2006.01); G06T 7/00 (2017.01); G06V 10/25 (2022.01)
CPC G06V 10/255 (2022.01) [A61B 5/441 (2013.01); G06T 7/0012 (2013.01); G06V 10/25 (2022.01); G06T 2207/30088 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A feature quantity extracting device, comprising:
at least one processor that executes a program stored in at least one memory, wherein the at least one processor is configured to:
acquire a captured image obtained by imaging a predetermined target;
specify a linear region extending in a plurality of mutually different directions, from a region of the captured image acquired by the at least one processor, the region including the predetermined target;
obtain pixel values in the specified linear region as one-dimensional data; and
extract a feature quantity from the linear region specified by the at least one processor by calculating a predetermined statistic value for the one-dimensional data,
wherein the at least one processor is configured to obtain the pixel values in the specified linear region as one-dimensional data of each of a first component, a second component, and a third component in a predetermined color space, and extract, as the feature quantity, at least one of 16 values that are a variance of the one-dimensional data of each of the first component and the second component, a gradient and a contribution ratio of a regression line of the one-dimensional data of each of the first component, the second component, and the third component, a variance of absolute values of differences among the one-dimensional data, gradients and contribution ratios of regression lines of the first component/the second component, and the first component/the third component among ratios of the one-dimensional data, and differences of maximum values and minimum values of the one-dimensional data.