Image analysis device, method for creating image feature information database, and design similarity determination apparatus and method

Patent metadata

Field	Value
Publication number	US-9659231-B2
Application number	US-201514878337-A
Country	US
Kind code	B2
Filing date	Oct 8, 2015
Priority date	Nov 28, 2014
Publication date	May 23, 2017
Grant date	May 23, 2017

Abstract

Provided is an image analysis device for specifying an image region of an object being the basis for design similarity determination. An image analysis device 100 includes units 112, 113 . The unit 112 moves original images Im 1 and Im 2 including a designated image region H with a rotationally symmetry on a first point P relative to a reference image Im 2 ( j ), calculates a correlation value cim(j) at each relative position, and detects the position of the point P in the images Im 1 , Im 2 , according to a geometric model, based on the amount of rotation θ(j) at which the cim(j) is the maximum and a vector v(j). The unit 113 specifies the range of the region H in the images Im 1 , Im 2 based on the position of the point P and the distribution of the brightness values of pixels in the images Im 1 , Im 2.

First claim

The invention claimed is:

1. An image analysis device, comprising:
at least one processor; and
at least one memory including computer program code,
wherein the at least one memory and the computer program code are configured to, with the at least one processor, cause the image analysis device to:
acquire an original image including a designated image region which is rotationally symmetric with respect to a first point;
generate a reference image by rotating the original image about a second point on the original image;
calculate a correlation value between the original image and the reference image at each relative position by virtually and relatively moving the original image with respect to the reference image;
detect the position of the first point in the original image, according to a geometric model indicating a relative positional relationship between the first point and the second point in the original image, on the basis of the amount of rotation of the reference image about the second point at a relative position between the original image and the reference image where the correlation value is the maximum and a vector indicating a deviation between the second point of the original image and the second point of the reference image; and
specify a range of the designated image region in the original image, based on the detected position of the first point and a distribution of brightness values of pixels in the original image.

2. The image analysis device according to claim 1 ,
wherein the at least one memory and the computer program code are further configured to, with the at least one processor, cause the image analysis device to generate a polar coordinate expansion image of the designated image region, which has the position of the first point as the origin, has an axis corresponding to a radius direction of the designated image region as a first axis, and has, as a second axis, an axis that corresponds to a circumferential direction of the designated image region and is perpendicular to the first axis, based on the first point and the range of the designated image region; and
analyze the polar coordinate expansion image and acquires feature information included in the polar coordinate expansion image.

3. The image analysis device according to claim 2 , wherein the at least one memory and the computer program code are further configured to, with the at least one processor, cause the image analysis device to:
generate an edge image of the original image and generate a polar coordinate expansion image of the edge image, which has a first point of the edge image as the origin, to generate the polar coordinate expansion image of the designated image region;
separate the polar coordinate expansion images into a plurality of segments which are partitioned by an edge line; and
compare the shape of each segment included in the polar coordinate expansion image with the shape of each segment included in an image obtained by inverting the polar coordinate expansion image in the direction of the second axis and determines whether each segment is symmetric with respect to a straight line parallel to the first axis.

4. The image analysis device according to claim 2 , wherein the at least one memory and the computer program code are further configured to, with the at least one processor, cause the image analysis device to:
generate an edge image of the original image and generate a polar coordinate expansion image of the edge image, which has a first point of the edge image as the origin, to generate the polar coordinate expansion image of the designated image region;
separate the polar coordinate expansion images into a plurality of segments which are partitioned by an edge line;
group a plurality of segments having a similar shape into one cluster; and
acquire a distance between the segments, which belong to each cluster and are adjacent to each other in the direction of the second axis, in the polar coordinate expansion image in the direction of the second axis.

5. The image analysis device according to claim 4 , wherein
on condition that the designated image region has a predetermined number of rotational symmetries which is equal to or greater than 2, the at least one memory and the computer program code are further configured to, with the at least one processor, cause the image analysis device to calculate an amount of deviation which is a difference between a theoretical distance between segments that are adjacent to each other in the direction of the second axis in each cluster in the second axis direction, and the actual distance between the adjacent segments in the second axis direction, which is acquired by the image analysis device, and
when the sum of the amounts of deviation in the one cluster is equal to or less than a predetermined threshold value, the at least one memory and the computer program code are further configured to, with the at least one processor, cause the image analysis device to estimate that the designated image region has the predetermined number of rotational symmetries.

6. The image analysis device according to claim 5 , wherein the at least one memory and the computer program code are further configured to, with the at least one processor, cause the image analysis device to:
calculate the amount of deviation between the adjacent segments belonging to each cluster for the predetermined number of rotational symmetries at each test distance obtained by multiplying a unit distance which is determined from the predetermined number of rotational symmetries by an integer equal to or greater than 1;
calculate, as the number of missing segments, a value obtained by subtracting 1 from the multiplied integer for the test distance having the minimum amount of deviation among the amounts of deviation; and
when the sum of the minimum amounts of deviation in one cluster is equal to or less than the threshold value and the sum of the numbers of missing segments in one cluster is the minimum, estimate that the designated image region has the predetermined number of rotational symmetries.

7. An apparatus that scores a similarity between a first designated image region included in a first image and a second designated image region included in a second image, and determines a similarity between designs included in the two images, comprising:
at least one processor; and
at least one memory including computer program code,
wherein the at least one memory and the computer program code are configured to, with the at least one processor, cause the image analysis device to
acquire an original image including a designated image region which is rotationally symmetric with respect to a first point,
generate a reference image by rotating the original image about a second point on the original image,
calculate a correlation value between the original image and the reference image at each relative position by virtually and relatively moving the original image with respect to the reference image,
detect the position of the first point in the original image, according to a geometric model indicating a relative positional relationship between the first point and the second point in the original image, on the basis of the amount of rotation of the reference image about the second point at a relative position between the original image and the reference image where the correlation value is the maximum and a vector indicating a deviation between the second point of the original image and the second point of the reference image,
specify a range of the designated image region in the original image, based on the detected position of the first point and a distribution of brightness values of pixels in the original image, and
perform scoring based on featur

Assignees

• Honda Motor Co Ltd

Inventors

• Sakai Saburo

Classifications

• G06T7/68 Primary

  of symmetry · CPC title

• G06V10/751

  Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching · CPC title

• G06V10/469

  Contour-based spatial representations, e.g. vector-coding · CPC title

• G06V10/759

  Region-based matching · CPC title

• G06K9/481

  Physics · mapped topic