Information processing system, information processing method, information processing device and control method and control program thereof, and communication terminal and control method and control program thereof
First Claim
1. An information processing system, comprising:
- a local feature storage which stores a piece of merchandise and m-number of first local feature which are respectively 1-dimensional to i-dimensional feature vectors and which are generated for each of m-number of local areas including each of m-number of feature points of an image of the merchandise after associating the piece of merchandise and m-number of first local feature with each other;
an imaging unit;
a first processor that executes the steps of;
extracting n-number of feature points from an image on a video captured by the imaging unit, andgenerating n-number of second local feature which are respectively 1-dimensional to j-dimensional feature vectors for n-number of local areas including each of the n-number of feature points; and
a second processor that executes the steps of;
selecting a smaller number of dimensions among the number of dimensions i of the feature vectors of the first local feature and the number of dimensions j of the feature vectors of the second local feature, andrecognizing that the merchandise exists in the image on the video when determination is made that a prescribed proportion or more of the m-number of first local feature which are feature vectors up to the selected number of dimensions correspond to the n-number of second local feature which are feature vectors up to the selected number of dimensions,wherein the first local feature and the second local feature are generated by dividing a local area including a feature point extracted from an image into a plurality of sub areas and generating feature vectors of a plurality of dimensions constituted by a gradient direction histogram in the plurality of sub areas, andwherein the first local feature and the second local feature are generated by deleting dimensions with a higher correlation between adjacent sub areas from the generated feature vectors of the plurality of dimensions.
1 Assignment
0 Petitions
Accused Products
Abstract
Merchandise management is implemented by recognizing a piece of merchandise in an image on a video in real time. A piece of merchandise and m-number of first local feature which are respectively 1-dimensional to i-dimensional feature vectors are stored after being associated with each other, n-number of feature points are extracted from an image on a video captured by an imaging unit, n-number of second local feature which are respectively 1-dimensional to j-dimensional feature vectors are generated, a smaller number of dimensions of the number of dimensions i and the number of dimensions j is selected, and a recognition that the merchandise exists in the image on the video is made when it is determined that a prescribed proportion or more of the m-number of first local feature of the selected number of dimensions correspond to the n-number of second local feature of the selected number of dimensions.
-
Citations
20 Claims
-
1. An information processing system, comprising:
-
a local feature storage which stores a piece of merchandise and m-number of first local feature which are respectively 1-dimensional to i-dimensional feature vectors and which are generated for each of m-number of local areas including each of m-number of feature points of an image of the merchandise after associating the piece of merchandise and m-number of first local feature with each other; an imaging unit; a first processor that executes the steps of; extracting n-number of feature points from an image on a video captured by the imaging unit, and generating n-number of second local feature which are respectively 1-dimensional to j-dimensional feature vectors for n-number of local areas including each of the n-number of feature points; and a second processor that executes the steps of; selecting a smaller number of dimensions among the number of dimensions i of the feature vectors of the first local feature and the number of dimensions j of the feature vectors of the second local feature, and recognizing that the merchandise exists in the image on the video when determination is made that a prescribed proportion or more of the m-number of first local feature which are feature vectors up to the selected number of dimensions correspond to the n-number of second local feature which are feature vectors up to the selected number of dimensions, wherein the first local feature and the second local feature are generated by dividing a local area including a feature point extracted from an image into a plurality of sub areas and generating feature vectors of a plurality of dimensions constituted by a gradient direction histogram in the plurality of sub areas, and wherein the first local feature and the second local feature are generated by deleting dimensions with a higher correlation between adjacent sub areas from the generated feature vectors of the plurality of dimensions. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19)
-
-
20. An information processing method in an information processing system including a local feature storage that stores a piece of merchandise and m-number of first local feature which are respectively 1-dimensional to i-dimensional feature vectors and which are generated for each of m-number of local areas including each of m-number of feature points of an image of the merchandise after associating the piece of merchandise and m-number of first local feature with each other,
the information processing method comprising: -
an imaging step of performing imaging; a second local feature generating step of extracting n-number of feature points from an image on a video captured in the imaging step and generating n-number of second local feature which are respectively 1-dimensional to j-dimensional feature vectors for n-number of local areas including each of the n-number of feature points; and a recognizing step of selecting a smaller number of dimensions among the number of dimensions i of the feature vectors of the first local feature and the number of dimensions j of the feature vectors of the second local feature, and recognizing that the merchandise exists in the image on the video when determination is made that a prescribed proportion or more of the m-number of first local feature which are feature vectors up to the selected number of dimensions correspond to the n-number of second local feature which are feature vectors up to the selected number of dimensions, wherein the first local feature and the second local feature are generated by dividing a local area including a feature point extracted from an image into a plurality of sub areas and generating feature vectors of a plurality of dimensions constituted by a gradient direction histogram in the plurality of sub areas, and wherein the first local feature and the second local feature are generated by deleting dimensions with a higher correlation between adjacent sub areas from the generated feature vectors of the plurality of dimensions.
-
Specification