Image-feature extraction method and computer-readable record medium with a program for making a computer execute steps of the method recorded therein

Claim 2

An image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel constituting a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; wherein the color space is a color space which is suited to human sensibility for colors and also is segmented into areas in units suited to human sensibility for colors.

Claim 3

The image-feature extraction method according to claim 1 , wherein said color space is a biconical HSV space which is segmented into a plurality of areas.

Claim 4

An image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; accumulating a number of pixels which belong to other areas in the number of pixels which belong to each of the areas according to color similarity between the areas; and generating a histogram of the color based on the number of pixels which belong to each of the areas after accumulation thereof.

Claim 5

An image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; accumulating a number of pixels which belong to other areas in the number of pixels which belong to each of the areas according to color similarity between the areas; and generating a histogram of the color based on the number of pixels which belong to each of the areas after accumulation thereof.

Claim 6

An image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; previously setting typical colors of the area; computing similarity between a color of each of the pixels and a typical color of each area; and determining a number of pixels which belong to each of the areas according to the computed similarity.

Claim 7

An image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; previously setting typical colors of the area; computing similarity between a color of each of the pixels and a typical color of each area; and determining a number of pixels which belong to each of the areas according to the computed similarity.

Claim 8

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of an object from the extracted object; setting points evenly spaced along an outline of the extracted object; computing the displacement of the angle between the lines that link the adjacent points; and extracting each computed displacement of the angles as feature parameters of the object.

Claim 9

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of an object from the extracted object; setting points evenly spaced along an outline of the extracted object; computing the displacement of the angle between the lines that link the adjacent points; and extracting each computed displacement of the angles as feature parameters of the object.

Claim 10

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object.

Claim 11

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object.

Claim 12

The image-feature extraction method according to claim 6 , wherein the previously specified area includes areas corresponding to the upper left and upper right corners of the image.

Claim 13

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object; wherein the feature parameters of the extracted object are extracted using the steps of, identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel constituting a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; wherein the color space is a color space which is suited to human sensibility for colors and also is segmented into areas in units suited to human sensibility for colors.

Claim 14

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object; wherein the feature parameters of the extracted object are extracted using the steps of, identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel constituting a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; wherein the color space is a color space which is suited to human sensibility for colors and also is segmented into areas in units suited to human sensibility for colors.

Claim 15

The image-feature extraction method according to claim 8 , wherein the previously specified area includes areas corresponding to the upper left and upper right corners of the image.

Claim 16

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object; wherein the feature parameters of the extracted object are extracted using the steps of, identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; accumulating a number of pixels which belong to other areas in the number of pixels which belong to each of the areas according to color similarity between the areas; and generating a histogram of the color based on the number of pixels which belong to each of the areas after accumulation thereof.

Claim 17

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object; wherein the feature parameters of the extracted object are extracted using the steps of, identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; accumulating a number of pixels which belong to other areas in the number of pixels which belong to each of the areas according to color similarity between the areas; and generating a histogram of the color based on the number of pixels which belong to each of the areas after accumulation thereof.

Claim 18

The image-feature extraction method according to claim 10 , wherein the previously specified area includes areas corresponding to the upper left and upper right corners of the image.

Claim 19

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object; wherein the feature parameters of the extracted object are extracted using the steps of, identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; previously setting typical colors of the area; computing similarity between a color of each of the pixels and a typical color of each area; and determining a number of pixels which belong to each of the areas according to the computed similarity.

Claim 20

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object; wherein the feature parameters of the extracted object are extracted using the steps of, identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; previously setting typical colors of the area; computing similarity between a color of each of the pixels and a typical color of each area; and determining a number of pixels which belong to each of the areas according to the computed similarity.

Claim 21

The image-feature extraction method according to claim 12 , wherein the previously specified area includes areas corresponding to the upper left and upper right corners of the image.

Claim 22

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object; wherein the feature parameters of the extracted object are extracted using the steps of, extracting an object from a prepared image; extracting feature parameters representing features of an object from the extracted object; setting points evenly spaced along an outline of the extracted object; computing the displacement of the angle between the lines that link the adjacent points; and extracting each computed displacement of the angles as feature parameters of the object.

Claim 23

An image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object; wherein the feature parameters of the extracted object are extracted using the steps of, extracting an object from a prepared image; extracting feature parameters representing features of an object from the extracted object; setting points evenly spaced along an outline of the extracted object; computing the displacement of the angle between the lines that link the adjacent points; and extracting each computed displacement of the angles as feature parameters of the object.

Claim 24

The image-feature extraction method according to claim 14 , wherein the previously specified area includes areas corresponding to the upper left and upper right corners of the image.

Claim 25

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; wherein the color space is a color space which is suited to human sensibility for colors and also is segmented into areas in units suited to human sensibility for colors.

Claim 26

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; wherein the color space is a color space which is suited to human sensibility for colors and also is segmented into areas in units suited to human sensibility for colors.

Claim 27

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; accumulating a number of pixels which belong to other areas in the number of pixels which belong to each of the areas according to color similarity between the areas; and generating a histogram of the color based on the number of pixels which belong to each of the areas after accumulation thereof.

Claim 28

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; accumulating a number of pixels which belong to other areas in the number of pixels which belong to each of the areas according to color similarity between the areas; and generating a histogram of the color based on the number of pixels which belong to each of the areas after accumulation thereof.

Claim 29

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; previously setting typical colors of the area; computing similarity between a color of each of the pixels and a typical color of each area; and determining a number of pixels which belong to each of the areas according to the computed similarity.

Claim 30

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel which forms a prepared image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; and extracting the generated color histogram as feature parameters representing features of the image; previously setting typical colors of the area; computing similarity between a color of each of the pixels and a typical color of each area; and determining a number of pixels which belong to each of the areas according to the computed similarity.

Claim 31

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of an object from the extracted object; setting points evenly spaced along an outline of the extracted object; computing the displacement of the angle between the lines that link the adjacent points; and extracting each computed displacement of the angles as feature parameters of the object.

Claim 32

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of an object from the extracted object; setting points evenly spaced along an outline of the extracted object; computing the displacement of the angle between the lines that link the adjacent points; and extracting each computed displacement of the angles as feature parameters of the object.

Claim 33

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object.

Claim 34

A computer-readable record medium with a program for making a computer execute the steps of the image-feature extraction method, which image-feature extraction method comprising the steps of: extracting an object from a prepared image; extracting feature parameters representing features of the object from the extracted object; extracting a color of each pixel which forms the image; clustering the extracted colors in a color space; identifying any cluster, of the clusters obtained in the clustering step, including a color appearing the most frequently in a previously specified area of the image as a cluster for a background area of the image; and extracting an area of the image formed with colors which belong to the clusters other than the identified cluster as the object.

Claim 36

An image data method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel constituting a first image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; extracting the generated color histogram as feature parameters for said first image; wherein the color space is a color space which is suited to human sensibility for colors and also is segmented into areas in units suited to human sensibility for colors; and comparing said feature parameters for said first image to feature parameters for a second image.

Claim 37

An image data method comprising the steps of: identifying to which area in a color space with a plurality of areas previously segmented thereinto each pixel belongs based on a color of each pixel constituting a first image; generating a color histogram of the image based on a number of pixels which belong to each of the areas; extracting the generated color histogram as feature parameters for said first image; wherein the color space is a color space which is suited to human sensibility for colors and also is segmented into areas in units suited to human sensibility for colors; and comparing said feature parameters for said first image to feature parameters for a second image.

Claim 38

The method of claim 22 , further comprising the step of storing said feature parameters for said second image in an image feature database.

Claim 39

The method of claim 23 , further comprising the step of retrieving said second image in response to said feature parameters for said second image being similar to said feature parameters for said first image.