KR101715655B1 - Image key certification method and system using color histogram and texture information of image - Google Patents

Abstract

Disclosed are a method and a system for image key authentication using color histogram and texture information of an image. A system for image key authentication according to an embodiment of the present invention comprises: an image input unit receiving an original image to be used as an image key of user authentication or an input image for user authentication; a feature extraction unit respectively extracting a first image feature value and a second image feature value through an image analysis with respect to the original image and the input image; a data storage unit storing the first image feature value in a database together with the original image; and an authentication performing unit which compares the second image feature value with the first image feature value stored in the data storage unit and determines whether to authenticate a user according to similarity.

Description

TECHNICAL FIELD The present invention relates to an image key authentication method and system using color histogram and texture information of an image,

The present invention relates to an image key authentication method and system using color histogram and texture information of an image.

With the development of the Internet, the number of networks connected to the Internet has increased rapidly, and the scope and content of information on the Internet has increased widely. Under such an Internet environment, various types of web services (e. G., E-commerce) as well as new types of web services have been developed to be applied to the Internet.

The web service provider system generally uses an ID and a password for identifying and authenticating a certain user for management and security reasons of the user. If you want a stronger authentication, you can use an encrypted public certificate, a security card (a random numbered array card), and a phone certificate. These methods can also help enhance authentication security when used in conjunction with each other.

As a related art, a system and a method for performing authentication only for a user registered in Korean Patent Laid-open No. 10-2015-0081387 are disclosed.

Korean Patent Publication No. 10-2015-0081387 (date of publication July 14, 2015) - User authentication system and method

Accordingly, the present invention provides a video key authentication method and system using color histogram and texture information of an image using color histogram and / or texture information of an image inherently possessed by all images as user authentication means.

The present invention can not restore the original image using only the color histogram and / or texture information of the image extracted and stored in the image source, and the thumbnail image provided as a hint to the authentication key for user convenience is used as the image key The present invention provides a video key authentication method and system using the color histogram and texture information of an image that can hardly restore 100% original video.

The present invention provides an image key authentication method and system using color histogram and texture information of an image to which a high-level security logic is applied, having elements that can be encrypted in various aspects from the viewpoint of database security.

Other objects of the present invention will become more apparent through the following preferred embodiments.

According to an aspect of the present invention, there is provided an image authentication method comprising: inputting an original image to be used as a video key of a user authentication or an input image for user authentication; A feature extraction unit for extracting a first image feature value and a second image feature value through an image analysis on the original image and the input image, respectively; A data storage unit for storing the first image feature value in a database together with the original image; And an authentication performing unit for comparing the second image feature value with the first image feature value stored in the data storage unit and determining whether the user is authenticated according to the similarity degree, wherein the first image feature value and the second image feature Values are feature vectors relating to the color histogram information of the original image and the input image, respectively, and the feature extracting unit includes three one-dimensional spaces divided on the basis of the elements constituting the color space, three two- And a color space selector for selecting a color space to analyze the color histogram among the seven color spaces including one three-dimensional space.

Wherein the feature extracting unit comprises: a differential quantizer that sets a bin region dividing interval for at least one of the color spaces to be smaller than another color space; A bin selection unit for selecting at least one of a type and a sequence of bins to analyze a color histogram among the bins; A weight applying unit for applying a predetermined weight to a frequency of a part of the blank areas among the blank areas divided for the color space; And an extraction region selection unit for selecting an extraction region to analyze the color histogram of the original image and the input image.

Wherein the feature extracting unit extracts the color histogram of the color histogram analyzed based on the result determined by at least one of the color space selecting unit, the differential quantizing unit, the bin selecting unit, the weight applying unit, And a feature vector generation unit for generating the feature vector by reflecting a spatial average position and a distribution of the pixels belonging to the feature vector.

According to another aspect of the present invention, there is provided a method of authenticating a user, comprising: (a) receiving an original image to be used as a video key of a user authentication; (b) extracting a first image feature value through image analysis on the original image; (c) the data storage unit storing the first image feature value in the database together with the original image; (d) receiving the input image for user authentication by the image input unit; (e) extracting a second image feature value through image analysis on the input image; (f) comparing the second image feature value with the first image feature value stored in the data storage unit and determining whether the user is authenticated according to the similarity, wherein the first image feature value and the second image feature value The second image feature value is a feature vector relating to color histogram information of the original image and the input image, respectively, and the step (b) includes three one-dimensional spaces classified on the basis of the elements constituting the color space, And a color space selection step of selecting a color space for analyzing the color histogram among seven color spaces including one three-dimensional space and one color space.

Wherein the step (b) comprises: a differential quantization step of setting a bin region dividing interval for one or more of the color spaces to be smaller than other color spaces; A bin selection step of selecting at least one of a type and an order of bin to analyze a color histogram among the bins; A weight applying step of applying a predetermined weight to a frequency of some of the blank areas among the blank areas divided for the color space; And an extraction region selection step of selecting an extraction region to analyze a color histogram of the original image and the input image, and the step (e) may perform the same sub-step as the step (b).

Wherein the step (b) and the step (e) are performed on the result determined by performing at least one of the color space selection step, the differential quantization step, the bin selection step, the weight applying step, and the extraction area selection step And generating the feature vector by reflecting the spatial average position and the distribution of the pixels belonging to the same bin with respect to the color histogram analyzed based on the color histogram.

Other aspects, features, and advantages will become apparent from the following drawings, claims, and detailed description of the invention.

According to the present invention, color histogram and / or texture information of an image inherently possessed by all images can be used as user authentication means.

In addition, the original image can not be restored only with color histogram and / or texture information of the image extracted and stored in the image source, and the thumbnail image provided as a hint to the authentication key for user convenience is used as the image key It is hard to restore 100% of the original image, so authentication security can be strengthened.

In addition, there are elements that can be encrypted in various aspects from the viewpoint of database security, so that a high level of security logic is applied.

1 is a block diagram of an image key authentication system using a color histogram and texture information of an image according to an embodiment of the present invention;
2 is a block diagram showing a detailed configuration of a feature extracting unit,
3 shows a three-dimensional color space,
4 shows a two-dimensional color space,
5 shows a one-dimensional color space,
6 is a diagram showing differential quantization,
7 is a diagram illustrating weighting application,
8 is a flowchart illustrating a method of authenticating a video key performed in a video key authentication system according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It is to be understood, however, that the invention is not to be limited to the specific embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terms first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, the terms "comprises" or "having" and the like refer to the presence of stated features, integers, steps, operations, elements, components, or combinations thereof, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

It is to be understood that the components of the embodiments described with reference to the drawings are not limited to the embodiments and may be embodied in other embodiments without departing from the spirit of the invention. It is to be understood that although the description is omitted, multiple embodiments may be implemented again in one integrated embodiment.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Reference will now be made in detail to the embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail.

In this specification, an image refers to a still image and may be represented by an image.

FIG. 1 is a block diagram of an image key authentication system using a color histogram and texture information of an image according to an embodiment of the present invention. FIG. 2 is a block diagram showing a detailed configuration of a feature extraction unit. FIG. 5 is a diagram illustrating a one-dimensional color space, FIG. 6 is a diagram illustrating differential quantization, and FIG. 7 is a diagram illustrating application of a weight. FIG.

The image key authentication system according to an exemplary embodiment of the present invention utilizes color histogram and / or texture information inherent in all images as a feature value for user authentication. The image intuitively conveys information and meaning to a person more than texts such as letters, numbers, and symbols. Thus, when a video is used as an authentication key, it may be more useful than a textual password in certain fields.

Referring to FIG. 1, the video key authentication system 100 includes an image input unit 110, a feature extraction unit 120, a data storage unit 140, and an authentication execution unit 150. The hint image providing unit 160, the encrypting unit 130, the decrypting unit 60, and the like may be further included.

The image input unit 110 may receive an original image (image key) for an image feature value to be stored in the database 50 and an input image (authentication image) for authentication. The input original image and input image may be a still image captured through a camera or a partial image captured from a moving image.

In the present embodiment, the original image may be stored in the database 50, and thumbnails of the original image may be stored together. Here, the thumbnail is a hint image provided for future user authentication. The thumbnail may be reduced in size only to the original image, or may be subjected to color modulation or morphological modification to prevent illegal use through capturing.

The feature extraction unit 120 extracts image feature values for the image (original image or input image) received from the image input unit 110. [ The extraction of the image feature value may be performed according to a predetermined or predetermined extraction criterion set by the user. The extraction criterion can be utilized as one of security enhancement logic in the user authentication process.

In the present embodiment, color histogram information or texture information may be utilized as the image feature value.

First, the color histogram information can be used to extract a histogram feature vector that shows a high degree of accuracy and a high degree of accuracy through various variables such as color space selection, differential quantization, bin selection, weighting, have.

Referring to FIG. 2, there is shown a detailed configuration of the feature extraction unit 120 for extraction of color histogram information.

The feature extraction unit 120 includes at least one of a color space selection unit 121, a differential quantization unit 122, a bin selection unit 123, a weight application unit 124, and an extraction region selection unit 125, And may further include a feature vector generation unit 126.

The color space selector 121 divides one image into seven color spaces, and selects a color space in which to analyze the color histogram.

Basically, the color is determined by the characteristics of the light reflected from the object. Based on the three primary colors designated for the purpose of standardization in the CIE, the color is defined as RGB color space, opponent color color space, YCbCr color space, HSV color space, Various color spaces are defined and used. In the present embodiment, it is assumed that each pixel of the image is placed in the RGB color space for the sake of understanding and explanation of the invention.

Each pixel of the image can be represented by a combination of red (R), green (G), and blue (B), and placed in seven color histogram spaces from three dimensions to one dimension based on the RGB values of each pixel .

If each pixel of the image basically has a YCbCr value, the RGB value may be obtained using the following equation, which is a conversion equation between the YCbCr value and the RGB value, and then placed in the seven color histogram spaces.

[Equation 1]

First, one pixel can be arranged in an RGB (Red, Green, Blue) space 200, which is a cubic space (three-dimensional space) in which X, Y, and Z axes respectively represent R, G, (See FIG. 3). Further, a GB (Green, Blue) space 212 corresponding to the YZ flat name in a square space (two-dimensional space) in which two of R, G and B are combined, a BR (Blue, Red) space 214 ) And the RG (Red, Green) space 216 corresponding to the XY plane (see FIG. 4). In addition, a R (Red) space 224 corresponding to the X axis, a G (Green) space 226 corresponding to the Y axis, and a G (Green) space 224 corresponding to the Y axis are linear spaces (one- And blue (B) spaces 222 corresponding to the Z axis (see FIG. 5).

One pixel includes an RGB space 200 as a three-dimensional space, a GB space 212 as a two-dimensional space, a BR space 214, an RG space 216, an R space 224 as a one-dimensional space, The color space 226, and the B space 222, respectively.

If all pixels belonging to the image are arranged in seven color spaces from one dimension to three dimensions based on RGB values, a color histogram for seven color spaces can be obtained.

In the present embodiment, the color space selector 121 selects a color space to be used for authentication among the color histograms of the seven color spaces. The selection of the color space may be based on user input or according to a predetermined algorithm according to an extraction criterion. In this case, what color space is selected can be used as security logic for authentication.

Referring again to FIG. 2, the differential quantization unit 122 performs a quantization process for dividing a bin, which is a unit for accumulating frequency numbers, into seven color spaces, (I.e., irregularly). As a method of discriminative quantization, for example, among the seven color spaces, there may be a method in which a color space to be emphasized has a smaller space division interval relatively to the other color space than the other color space.

Referring to FIG. 6, differential quantization is applied in the process of performing quantization on three one-dimensional color spaces.

The B space 222 and the R space 224 are divided into eight empty regions, but the G space 226 is divided into 16 empty regions. That is, the G space 226 is more finely divided than the B space 222 or the R space 224, so that green can be more finely distinguished than blue or red.

In the case of applying the differential quantization, the interval between the bins is made irregular rather than being constant, so that the bin can be used as the security logic itself.

The total number of beans can be reduced compared to delineating beans for every color space, thereby reducing the computation time in the authentication process, thereby enabling quick authentication. Also, at the same time, the accuracy of authentication can be maximized by relatively finely dividing the color space to emphasize accuracy.

Referring back to FIG. 2, when an image is divided into a plurality of bins through quantization for seven color spaces, the bin selecting unit 123 can select a bin to be used for authentication. All of the bins may be selected, and one or more bins for user authentication may be arbitrarily selected, and the feature vectors may be extracted based on the frequencies belonging to the bins by limiting the selected bins to the selected bins. In this case, which bin is selected can be used as security logic in authentication.

When a plurality of bins are selected, the bin selection unit 123 may determine the order in which the selected plurality of bins are used or stored according to a user's input or according to a predetermined algorithm. In this case, the selected sequence of beans can also be used as security logic for authentication.

In addition, the weight applying unit 124 may select a bin to be emphasized among the bins divided through the quantization, and weight the frequency of the bin. Here, the weight reflects the degree of interest of the color gamut.

Referring to FIG. 7A, among the 16 bins divided for the G space 226, a predetermined weight is applied only to four bins whose G value is close to 255, Quot; value ", it is possible to obtain a high score.

Referring to FIG. 7 (b), by applying a predetermined weight only to bins whose G value is close to 255 and whose R value and B value are close to 0 out of a plurality of bins divided for the RGB space 200, And the blue component can be given a high score for an image having a small number of pixels with a large green component.

The extraction region selection unit 125 selects an area from which to extract the feature vector by the color histogram. A color histogram can be constructed only for a part of the image but not for the whole area, so that the feature vector can be extracted. In this case, which area is selected as the extraction area can be used as security logic in authentication.

Referring back to FIG. 2, the feature vector generation unit 128 generates a feature vector based on an extraction criterion, such as a color space selection unit 121, a differential quantization unit 122, a bin selection unit 123, a weight application unit 124, The color histogram vector can be extracted by finding the bin matching each pixel of the image based on the result determined by the region selection unit 125 or the like and calculating the matching frequency of each of the bins for all the pixels as the frequency. Then, a feature vector for a spatial average position and a distribution can be generated with respect to pixels belonging to the same bin.

The feature extraction unit 120 extracts feature vectors from the extracted feature vectors based on extraction criteria such as color space selection, differential quantization, bin selection, weight application, and extraction region selection as described above A preprocessing process such as brightness correction may be performed.

In addition, the feature extraction unit 120 may generate a feature vector based on the texture information in addition to the feature vector based on the color histogram.

The texture information is a texture feature value described as periodicity or half periodicity between pixels (pixels) included in the image. By dividing the frequency domain in each direction and radial direction in the radon region through a human visual filter A texture feature vector may be extracted. The texture feature vector includes elements related to the perception characteristics of human texture such as regularity, directionality, and ratio, and can be used for the similarity analysis between the original image and the input image for user authentication.

Referring back to FIG. 1, when the image input from the image input unit 110 is an original image to be used for user authentication, the data storage unit 140 stores the original image, which is extracted from the feature extraction unit 120, And stores them in the database 50 as user authentication data.

The user authentication data stored in the data storage unit 140 may be encrypted by the encryption unit 130 at the previous stage and stored in the database 50. [ One of various encryption algorithms related to data encryption can be applied to the encryption scheme, and a detailed description will be omitted for those skilled in the art.

When the image input from the image input unit 110 is an input image for user authentication, the authentication performing unit 150 searches the database 50 for the second image feature value extracted by the feature extracting unit 120 for the input image And compares the first image feature value with the first image feature value of the original image included in the one user authentication data to determine whether or not to match and determine whether to authenticate.

When the user authentication data stored in the database 50 is encrypted by the encryption unit 130, the second image feature value is also encrypted in the encryption unit 130, and then the first image feature value and the second image feature value A comparison can be made to the value.

If the input image is not identical to the original 100%, that is, if the input image is similar to the original, the second image feature value may be encrypted instead of encrypting the first image feature of the user authentication data stored in the database 50 Values may be decoded by the decoding unit 60 and the comparison may be performed. This is because some data may be lost in the process of decrypting the encrypted user authentication data. If the similarity of about 80-90% is required, the decryption may be utilized.

In addition, the video key authentication system 100 may further include a hint image providing unit 160 for providing (e.g., displaying) a hint image for giving a hint about a video key to a user who wishes to perform user authentication.

The hint image is a thumbnail for the original image, and it can help the user to input the image for authentication by allowing the user to confirm what the original image is registered in advance for image authentication.

It is difficult to restore 100% of the original image used as a video key by thumbnail alone. However, in this case, there is a concern that the thumbnail, which is a hint image, may be reduced due to the camera shots or screen captures in the hint image providing process. In addition to the size reduction of the original image, It is possible to raise the security so that the image characteristic value can be changed in the case of illegal use so that it can not be used for authentication.

In this embodiment, the extraction criterion may be arbitrarily selected in the system. Or may be selected by logic created using additional authentication (e.g., a four-digit password) with the video key.

FIG. 8 is a flowchart illustrating an image key authentication method performed in an image key authentication system according to an embodiment of the present invention.

First, the user authentication data registration process performed in the video key authentication system 100 is as follows.

The image input unit 110 receives an original image (image key) to be used for user authentication (step S310).

The feature extraction unit 120 analyzes the original image and extracts a first feature value of the color histogram information and / or the texture information (step S320).

The data storage unit 140 stores the extracted first image feature value together with the original image as user authentication data in the database 50 (step S330). In the process of storing in the database 50, the first image feature value may be encrypted according to a predetermined encryption algorithm and stored in an encrypted state.

Next, the user authentication process performed in the video key authentication system 100 is as follows.

The hint image providing unit 160 may provide a hint image of the original image registered for image security to the user who wants to log in (step S340). The hint image may be a thumbnail that is reduced in size with respect to the original image, and may be provided with some color or shape modulated to prevent illegal use by shooting or capturing a hint image.

The image input unit 110 receives an input image for image authentication, that is, an authentication image, from a user who wants to log in (step S350).

The feature extraction unit 120 analyzes the input image and extracts a second image feature value for color histogram information and / or texture information (step S360). Here, the extraction of the second image feature value may be performed based on an extraction criterion applied when extracting the first image feature value for the original image pre-registered for the user.

The authentication performing unit 150 compares the first image feature value of the user authentication data stored in the database 50 with the second image feature value extracted in step S360 (step S370), and based on the comparison result, Authentication is performed (step S380).

If the input image is required to be the same as the original image, the authentication is determined according to whether the first image feature value and the second image feature value are 100% identical. If the first image feature value is encrypted, the second image feature value may be encrypted and then compared with the second image feature value.

If the input image is required to have a degree of similarity (for example, about 80-90% similarity) to the original image, the first image feature value is decrypted and then compared with the second image feature value have. Encryption / decryption is used to protect the user authentication data stored in the database 50 when there is no need for 100% matching of the image authentication .

The video key authentication method according to the present invention can be implemented as a computer-readable code on a computer-readable recording medium. The computer-readable recording medium includes all kinds of recording media storing data that can be decoded by a computer system. For example, it may be a ROM (Read Only Memory), a RAM (Random Access Memory), a magnetic tape, a magnetic disk, a flash memory, an optical data storage device, or the like. In addition, the computer-readable recording medium may be distributed and executed in a computer system connected to a computer network, and may be stored and executed as a code readable in a distributed manner.

The video key authentication system and method according to the present embodiment can be applied to the following use examples.

According to the first use example, it can be used for authentication at the time of login of a web site using a photograph.

Generally, websites use user IDs and passwords to authenticate users. If you want a stronger authentication, you can use an encrypted certificate and a security card (a random numbered array card). In addition, if a photograph of the image key authentication method according to the present embodiment is used as an authentication key, authentication can be performed to make it easier and more interesting.

You can also specify a user's permission limit in authentication via photo. For example, if you have an image of the original, you can maintain the master privilege because it is 100% identical. If you have a picture that is slightly blurred or slightly different from the original, it will not be 100% on a color histogram basis, even if the picture is similar to the original. In this case, the authority of the authenticator can be different according to the difference from the original (%).

According to the second usage example, the user authentication system using a photograph can be utilized as a locker in a hotel room.

The hotel room is equipped with lockers for valuables of travelers. Currently, lockers are managed by simply using numeric passwords. This is because the length of stay for guests is short. That is, it is inconvenient for users to generate IDs for issuance of authorized certificates, or to provide a security card (a random number array card).

However, there are travelers who feel uneasy about security just by using a password (typically four digits). Therefore, if a passenger uses a photo in his or her smartphone as an authentication key together with a password, it will provide a stronger authentication but also a convenient and amusing authentication method.

According to the third usage example, it can be utilized when sending and receiving a text message (including an interactive text message such as a message of KakaoTalk), or when a photo is authenticated on a bulletin board.

When secretly sending and receiving texts between lovers, it is possible to authenticate with the pictures they share with each other so that only authenticators can send and receive texts. In addition, if you want to post a bulletin on a bulletin board (for example, a group bulletin board such as a line band), and send a picture used as an authentication key to a corresponding person, only the person can view the bulletin board, .

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the invention as defined in the appended claims. It will be understood that various modifications and changes may be made.

100: video key authentication system 110: video input unit
120: Feature extraction unit 130:
140: Data storage unit 150: Authentication performing unit
160: hint image providing unit 50: database
60:
121: Color space selection unit 122: Differential quantization unit
123: bin selection unit 124: weight application unit
125: extraction region selection unit 126: feature vector generation unit

Claims (7)

An image input unit for receiving an original image to be used as a video key for user authentication or an input image for user authentication;
A feature extraction unit for extracting a first image feature value and a second image feature value through an image analysis on the original image and the input image, respectively;
A data storage unit for storing the first image feature value in a database together with the original image; And
And an authentication unit for comparing the second image feature value with the first image feature value stored in the data storage unit and determining whether the user is authenticated according to the similarity,
Wherein the first image feature value and the second image feature value are feature vectors related to color histogram information on the original image and the input image, respectively,
Wherein the feature extracting unit extracts, from among three color spaces, three three-dimensional spaces, three one-dimensional spaces, and three color spaces, which are classified based on the elements constituting the color space, And a color space selection unit for selecting the video key.
The method according to claim 1,
The feature extraction unit may extract,
A differential quantizer configured to set a bin region division interval smaller than other color spaces for at least one of the color spaces;
A bin selection unit for selecting at least one of a type and a sequence of bins to analyze a color histogram among the bins;
A weight applying unit for applying a predetermined weight to a frequency of a part of the blank areas among the blank areas divided for the color space;
Further comprising at least one of an original image and an extraction area selection unit for selecting an extraction area for analyzing a color histogram of the input image.
3. The method of claim 2,
Wherein the feature extracting unit extracts the color histogram of the color histogram analyzed based on the result determined by at least one of the color space selecting unit, the differential quantizing unit, the bin selecting unit, the weight applying unit, Further comprising a feature vector generation unit configured to generate the feature vector by reflecting a spatial average position and a distribution of pixels belonging to the feature vector.
(a) receiving an original image to be used as a video key of a user authentication;
(b) extracting a first image feature value through image analysis on the original image;
(c) the data storage unit storing the first image feature value in the database together with the original image;
(d) receiving the input image for user authentication by the image input unit;
(e) extracting a second image feature value through image analysis on the input image;
(f) comparing the second image feature value with the first image feature value stored in the data storage unit and determining whether the user authentication is performed according to the similarity,
Wherein the first image feature value and the second image feature value are feature vectors related to color histogram information on the original image and the input image, respectively,
In the step (b), the feature extracting unit extracts, from seven color spaces including three one-dimensional spaces, three two-dimensional spaces, and one three-dimensional space, which are classified based on the elements constituting the color space, And a color space selection step of selecting a color space to analyze the histogram.
5. The method of claim 4,
The step (b)
A differential quantization step of setting a bin region dividing interval for at least one of the color spaces to be smaller than other color spaces;
A bin selection step of selecting at least one of a type and an order of bin to analyze a color histogram among the bins;
A weight applying step of applying a predetermined weight to a frequency of some of the blank areas among the blank areas divided for the color space;
And an extraction region selection step of selecting an extraction region to analyze a color histogram of the original image and the input image,
Wherein the step (e) performs the same sub-steps as the step (b).
6. The method of claim 5,
Wherein the step (b) and the step (e) are performed on the result determined by performing at least one of the color space selection step, the differential quantization step, the bin selection step, the weight applying step, and the extraction area selection step And generating a feature vector by reflecting a spatial average position and a distribution of pixels belonging to the same bin with respect to the color histogram analyzed based on the analyzed histogram.
A computer-readable recording medium on which a program for carrying out the method according to any one of claims 4 to 6 is recorded.

Images