JPH1153386A - System and method for image retrieval and image retrieval program recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily retrieve an image that is same as or similar to the image to be retrieved out of plural images and to obtain the objective and quantitative retrieval result by storing previously the feature values of plural object images having approximately the similar shapes. SOLUTION: When a user operates a photographing button on a control panel, a control part 11 turns on an illumination instrument 4 via an illumination part 15 and photographs the face photo of the user by a CCD camera, etc., via a photographing part 14. Then the part 11 extracts the measurement value of face parts such as the width, height, angles, space, etc., of eyes as the feature values. Based on these measurement value, the corresponding face parts images are selected from a face parts data base 16 and synthesized with the image data on the background, etc., which are inputted by the user. Thus, a portrait image is produced. Then the part 11 sends a seal form to an output part 19 from a store part 17 via a carrier part 18 and prints the portrait image on the seal form via the part 19.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image retrieval system for retrieving an image that is the same as or similar to a retrieval target image from a plurality of target images based on a characteristic amount representing the characteristic of the target. The present invention relates to a method and an image search program recording medium.

[0002]

2. Description of the Related Art In recent years, not only mass media such as televisions and newspapers but also OA devices such as personal computers have frequently used face images for identification of persons. For example, face image data captured by a camera or face image data read from a face photograph by a scanner or the like is input to a personal computer, and a face image subjected to predetermined image processing is partially inserted to create a document or the like. May be.
In some cases, it is used for recreational purposes such as creating a seal or stamp by combining an appropriate background or the like with a captured face image.

On the other hand, as a process used for authentication of a person or criminal investigation in a security system, it is conceivable to search for the same face image as a search target in a database storing a plurality of face images in advance. like this,
The face image search process can also be used for recreational purposes such as searching for a person who looks just like a celebrity.

[0004]

However, in a conventional apparatus for processing a face image, although a plurality of face images can be stored in a database, a face image to be searched from the plurality of face images stored in the database can be obtained. There is no efficient means for retrieving the same or similar face image as that of the original. Therefore, there is no choice but to rely on the judgment by the searcher's visual recognition, and not only does the retrieval process become complicated, There is a problem that it is not possible to perform a search based on accurate and quantitative judgment.

[0005] The above problem is caused not only in processing of a face image,
The same applies to the processing of any article such as agricultural products or industrial products.

SUMMARY OF THE INVENTION It is an object of the present invention to be able to easily retrieve an image identical or similar to an image to be searched from a plurality of images based on a feature amount extracted from the image of the object, and to objectively and quantitatively determine the image. It is an object of the present invention to provide an image search system, an image search method, and an image search program recording medium capable of obtaining a basic search result.

[0007]

According to the first, eighth and fifteenth aspects of the present invention, the feature amounts of a plurality of objects having substantially similar shapes are stored in a feature amount storage unit, and the search target image is stored. An image which can be determined to be the same as or similar to the image to be searched based on the feature amount of (i) is searched in the feature amount storage unit.

In the inventions described in claims 1, 8 and 15, an image having the same or similar feature as the image to be searched is searched from among the plurality of images stored in the feature storage. Therefore, the identity and similarity of the images are objectively and quantitatively determined based on the feature amounts of the image of the target object such as dimensions and angles.

[0009] The invention described in claims 2, 9 and 16 is:
Receiving an input of an image of the target object, extracting a feature amount from the input image of the target object, and searching the feature amount storage unit for an image having the same or similar feature amount of the extracted image of the target object. Features.

According to the second, ninth and sixteenth aspects, a feature amount is extracted from the input image of the object, and an image identical or similar to the extracted feature amount is searched in the feature amount storage unit. You. Therefore, the identity and similarity between the input image of the target object and the known image stored in advance in the feature amount storage unit are objectively and quantitatively determined.
The invention according to Claims 3, 10 and 17 receives an image of an image of a target object, extracts a characteristic amount from the input image of the target object, and calculates a characteristic amount of the extracted image of the target object. The image processing apparatus is characterized in that an image having the same or similar feature amount as a preset feature amount of the reference image is retrieved from the feature amount storage unit while being stored in the amount storage unit.

According to the third, tenth, and seventeenth aspects of the present invention, the feature amount extracted from the input image of the object is stored in the feature amount storage unit, and is the same as or similar to the preset reference image. An image of the feature amount is searched in the feature amount storage unit. Therefore, the identity and similarity of the input plurality of target images with the reference image are objectively and quantitatively determined.

According to a fourth aspect of the present invention, there is provided a server for storing a feature amount of an image of an object and retrieving an image having the same or similar feature amount as a preset feature amount of a reference image. Transmitting and receiving feature amount data and search data between a plurality of terminal devices each receiving an image of the image of the object and extracting a feature amount from the input image of the object. I do.

[0013] In the inventions described in claims 4, 11 and 18, images of a large number of objects are input from a plurality of terminal devices. Accordingly, the hit rate in searching for an image that is the same as or similar to the known image or the reference image is improved.

The invention described in claims 5, 12 and 19 is characterized in that the feature amount is a combination of feature amounts individually extracted for each component of the image of the object.

[0015] In the inventions described in claims 5, 12 and 19, the same or similar image is determined based on a combination of characteristic amounts such as dimensions and angles individually extracted for each component of the object. . Therefore, the identity and similarity of the image of the target object are determined as the sum of the determination results of the identity and similarity of each of the constituent parts of the target object, and the accuracy of the determination result is improved.

According to a sixth aspect of the present invention, a plurality of image component data classified according to the characteristic amount for each component of the image of the object are stored in the component storage unit in advance, and The image component data corresponding to the feature value of each component of the image is selected from the component storage unit as the feature value.

According to the inventions described in claims 6, 13 and 20, image component data corresponding to dimensions, angles and the like extracted individually for each of the constituent parts of the object are selected, and a set of a plurality of image component data is selected. The identity and similarity of the image of the target object are determined based on the feature amount. Therefore, the determination of the identity and similarity of the image of the target object is simplified based on the classified image component data of each component.

According to a seventh aspect of the present invention, in the part storage section, portrait image data is stored as a plurality of image part data classified for each of the constituent parts of the face according to the characteristic amount. It is characterized by.

According to the invention described in claims 7, 14 and 21, the face part data of the portrait corresponding to the dimensions and angles individually extracted for each of the face components such as the contour, eyes, nose and mouth are provided. The identity and similarity of the face images are determined based on the selected portrait image, which is a set of face part data. Therefore, the entertaining property in determining the identity and similarity of the face images is improved.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to an example in which the present invention is applied to a portrait creating apparatus for creating a sticker on which a portrait of a user's face image is printed.

FIG. 1 is an external view of a portrait creating apparatus to which an embodiment of a face image search system, a face image search method and a face image search program storage medium according to an embodiment of the present invention is applied. In the portrait creating apparatus 1, a window 3 is formed in the upper part of the front of the housing 2, and lighting devices 4 such as fluorescent lamps are arranged on the left and right of the window 3. An operation panel 5 is provided substantially horizontally below the window 3 on the front of the housing 2. The operation panel 5 includes a push button 5a.
And a lever 5b. A coin insertion slot 6a and a coin return slot 6b of the coin processing unit 6 are exposed below the operation panel 5 on the front of the housing 2, and an outlet 7 is formed further below the coin insertion slot 6a. FIG. 2 is a block diagram showing a configuration of the portrait creating apparatus 1. The portrait creation device 1 includes a control unit 11 configured by a microcomputer or the like, a coin processing unit 6, an operation unit 12, a display unit 13, an imaging unit 14, an illumination unit 15, a face part database 1
6, storage unit 17, transport unit 18, output unit 19, and communication unit 2
0 and the like. Control unit 1
1 is a program previously written in the memory or C
The input / output devices are generally controlled according to a program supplied from a storage medium such as a DROM. Coin processing unit 6
Accepts the insertion of a coin by the user at the coin insertion slot 6a, performs processing of discrimination of the inserted coin, identification of authenticity, calculation of change, and the like.
Return at b.

The operation unit 12 is provided with an operation panel 5 by a user.
The input of setting data for creating a seal or the like by operating the push button 5a or the lever 5b arranged on the device is received.
The display unit 13 displays, to the user, an operation method related to a creation operation of a seal or the like and a progress status of the operation via a display such as a CRT exposed to the front of the housing 2 through the window unit 3.
The imaging unit 14 is configured by a CCD camera or the like,
A face image of the user is captured through the window 3. Lighting unit 15
When the imaging unit 14 captures a face image, the lighting unit 4 arranged on the left and right of the window 3 is turned on to illuminate the user's face.

As shown in FIG. 3, the face part database 16 stores a plurality of face part images for creating a portrait image for each face part of eyes, nose, mouth, eyebrows, ears, contours, and hair. I have. More specifically, the face part database 16 stores, for example, the width of the eyes,
Based on measured values such as height, angle, and interval, a face part image corresponding to each stage when a change in the size or shape of each facial part is divided into a plurality of stages is stored. Each face part image is assigned an individual face part code, and each face part image can be specified by this face part code.

By providing the face part database 16 for each of a plurality of image types and for each of a plurality of facial expressions, any one of the image types and facial expressions can be selected to create a portrait image. Is also good.

The storage section 17 stores sticker sheets. The transport unit 18 transports the sticker sheet stored in the storage unit 17 to the outlet 7 via the output unit 19. The output unit 19 prints a portrait of the face image on the sticker paper transported by the transport unit 18. In addition, the portrait creating apparatus 1
A stamp of a portrait of a user's face image may be created. In this case, the stamp material is stored in the storage unit 17, and a process such as printing is performed on the stamp material transported by the transport unit 18 in the output unit 19. The communication unit 20 transmits and receives data between the control unit 11 and a server described later.

FIG. 4 is a diagram showing a configuration of a face image search system including the portrait drawing apparatus. The face image search system 31 is configured by connecting a plurality of portrait creating apparatuses 1 to a server 32. Each portrait creation device 1 is connected via a data communication line such as a dedicated line or a public telephone line network. The server 32 includes a control unit 42 including a database 41, a registration engine 42a, and a search engine 42b,
And a communication unit 43. Control unit 4
The second registration engine 42a performs a process of writing data to the database 41, and the search engine 42b performs a process of searching data in the database 41.

The database 41 is, for example, shown in FIGS.
It is composed of a registered user database 52, a first reference face image database 53, a second reference face image database 54, and a same face image database 55 shown in FIG. The registered user database 52 shown in FIG. 5 includes a user code 52a and a name 52b for specifying the user, and a face image captured by the imaging unit 14 of the portrait creation device 1 for each user who creates a portrait in each portrait creation device 1. A face image data code 52c for specifying data,
Permission data 52d indicating whether to permit the disclosure of own face image data to others, a face part code 52e for specifying a face part image selected based on the face image data, and a user such as age and gender. Each personal information 52f is stored.

In order to configure the database 52, the portrait creating apparatus 1 performs processing such as the name 52b, the permission data 52d, the personal information 52f, and the like after the creation of the portrait by the user or during the creation as described later. Of user input operations.

The first reference face image database 53 shown in FIG. 6 stores combinations of face part images constituting a beautiful face in order to search for a user corresponding to a predetermined beautiful face. That is, the database 53 stores, for each of the plurality of beautiful faces, a beautiful face code 5 for identifying each beautiful face.
3a, a face part code 53b of a face part image to be selected in each face part, a face image data code 53c for specifying image data of a beautiful face, and a user code 53d of a registered user corresponding to a beautiful face are stored. I do.

The second reference face image database 54 shown in FIG. 7 stores a combination of face part images when creating a portrait of a celebrity face image in order to search for a user similar to a celebrity. That is, the database 54 specifies, for each of the plurality of celebrities, the name and code 54a of the celebrity for specifying the celebrity, the face part code 54b of the face part image to be selected in each face part, and the face image data of the celebrity. Image data code 54c for registration and a user code 54d of a registered user having a face part code according to the same combination as the celebrity face part code 54b
Is stored.

It should be noted that not only the case where all face parts completely match the celebrity face part code, but also the registered user's user code which approximates the celebrity face part code for all face parts, and some face parts The user code of the registered user that matches the celebrity face part code for the part and approximates the same celebrity face part code for other face parts may also be stored.

The same face image database 55 shown in FIG.
Searches for registered users with similar face images,
A corresponding registered user code is stored for each combination of face part images. That is, the database 55 stores the combination 55a of the face part code and the user code 55 of the registered user whose combination of the face part code matches each combination.
b is stored.

It should be noted that not only a case where all face parts completely match the face part code of another person, but also a user code of a registered user who is similar to the face part code of another person for all face parts, and some face parts The user code of the registered user that matches the face part code of another person with respect to the part and may be similar to the face part code of the other person with respect to other face parts may be stored.

However, the database 55 is not essential,
By searching for registered face images similar to each other with reference to the registered user database 52 shown in FIG. 5,
The capacity of the database 51 may be reduced by making the database 55 unnecessary.

FIG. 9 is a flowchart showing a processing procedure of the control unit of the portrait drawing apparatus. Portrait drawing device 1
After the power is turned on, the control unit 11 waits for a user to insert a coin (s1). When the user inserts a coin, the coin processing unit 6 performs coin processing such as authenticity discrimination, denomination discrimination, and change processing (s2). ), When the coin processing is properly completed (s3), a screen for instructing photography is displayed on the display (s4). When the user operates the shooting button on the operation panel 5 (s5), the control unit 11 turns on the lighting device 4 via the lighting unit 15, and captures a face image of the user using a CCD camera or the like via the imaging unit 14 ( s6).

Next, the control unit 11 displays a screen for instructing input of image data on the display via the display unit 13 (s7), and displays images of the portrait, expression, background, etc. of the portrait on the operation panel 5 of the user. The input operation of the application data is received (s8). The image data includes the shape, position, and characters of the glasses that are selectively input. After that, the control unit 11 creates a portrait image after waiting for an input operation of the setting completion by the user (s9, s10).

In the process of creating a portrait image, the control unit 1
1 extracts, for each face part, measured values such as eye width, height, angle, and interval as feature amounts from the captured face image of the user and extracts a corresponding face part image based on the measured values. Is selected from the facial part database 16 and the image data such as the background input by the user is synthesized to create a portrait image. Thereafter, the control unit 11 transports the sticker sheet from the storage unit 17 to the output unit 19 via the transport unit 18, and prints a portrait image on the seal sheet via the output unit 19.

During or after the portrait image creation process, the control unit 11 accepts the user's selection of whether to execute the data registration process and various search processes (s11, s12). . When the user selects the execution of the data registration process, the control unit 11 displays a screen for instructing input of the registration data on the display via the display unit 13 (s13), and displays the name of the user and the portrait image. Accepts input of registration data such as disclosure availability, age and gender. Further, the control unit 11 communicates with the server 32 via the communication unit 20, and transmits the registration data to the server 32 together with the face part image code (s14).

When the user selects the execution of the search process, the control unit 11 displays a screen for instructing the input of the search content on the display via the display unit 13 (s15), and determines whether the user is a beautiful face. A selection input of a search content to determine whether to search for a celebrity similar to the user or a registered user similar to the user is received (s16). When the user performs a selection input of the search content, the control unit 11 communicates with the server 32 via the communication unit 20. When the user has completed the registration process, the control unit 11 transmits the user code. If the search is not completed, the user code and the face part image code are transmitted to the server 32 (s17). When the search result transmitted from the server 32 is received (s18), the search result is displayed via the display unit 13. Is displayed on the display (s19). Control unit 11
When the printing process of the sticker in the output unit 19 is completed, the sticker on which the portrait image has been printed is conveyed to the outlet 7 via the conveying unit 18 (s20, s21).

FIG. 10 is a flowchart showing a processing procedure of the control unit of the server. Control unit 42 of server 32
Are waiting for the communication unit 43 to receive the data transmitted from the portrait drawing apparatus 1 (s31, s32),
When the communication unit 43 receives the registration data transmitted from the portrait drawing apparatus 1, the registration engine 42a in the control unit 42
Stores the data constituting the registration data in the database 41
Is written into the registered user database 52 (s3).
3).

On the other hand, when the communication section 44 receives the search data transmitted from the portrait drawing apparatus 1, the search engine 42b in the control section 42 performs the first search based on the data specifying the search content included in the search data. In any one of the reference face image database 53, the second reference face image database 54, and the same face image database 55, based on a combination of face part codes included in the search data or a user code included in the search data. The same combination as the combination of the face part codes read from the user database 16 is searched (s34). Search engine 42b
Transmits the search result to the portrait creating apparatus 1 via the communication unit 43 (s35). The search result transmitted from the server 32 is received by the portrait creating apparatus 1 in s18 described above.

For example, when search data of a beautiful face is transmitted from the portrait drawing apparatus 1, the search engine 42b of the control unit 42 determines the combination of the user's face part codes included in the search data in the first reference face image database. Search at 53. When the combination of the face part codes of the user matches any of the combinations of the face part codes of the beautiful face stored in the first reference face image database 53, the search engine 42b extracts the user code. The search result is stored in the corresponding area of the reference face image database 53 and the search result indicating that the user is a beautiful face is transmitted to the portrait creating apparatus 1 via the communication unit 43.

When the celebrity search data is transmitted from the portrait drawing apparatus 1, the search engine 4 of the control unit 42
2b searches the second reference face image database 54 for a combination of the user's face part codes included in the search data. If the combination of the user's face part codes matches any of the combinations of celebrity face part codes stored in the second reference face image database 54, the search engine 42b sets the user code to the second reference face image database. Is stored in the corresponding area of the reference face image database 54, and a search result indicating that the user is similar to a specific celebrity is transmitted to the portrait creating apparatus 1 via the communication unit 43.

Further, when search data of another registered user similar to the user is transmitted from the portrait drawing apparatus 1, the search engine 42b of the control unit 42 sets the combination of the face part codes of the user included in the search data to the same. A search is performed in the face image database 55, and the user code included in the search data is stored in a corresponding area of the same face image database 55. When there is another registered user code having the combination of the face part codes in the same face image database 55, the search engine 42b reads out the face image data of another registered user similar to the user from the database 16. The search result including the face image data is transmitted to the communication unit 4.
3 to the portrait drawing apparatus 1.

With the above processing, in the face image retrieval system according to this embodiment, the face image of the user is combined with the face image selected based on the measured value of each face part extracted from the user's face image. Reference images of the same or similar beautiful faces, celebrities, etc. and face images of other registered users,
You can search quickly with simple processing. In addition, since the search is basically performed based on the measurement values of each face part extracted from the face image, an objective and quantitative search result can be obtained. Furthermore, since the portrait image creation system 1 is connected to the server 32 to form a face image search system, a large number of users can search for the same or similar users, and the hit rate in the search results can be reduced. Can be improved.

By providing the database 41 and the control unit 42 of the server 32 in the portrait creating apparatus 1,
Although the number of users to be searched decreases, the processing time can be shortened by eliminating the communication time with the server 32.

The face part database 16 is stored in the server 3
2, the configuration of each portrait creating apparatus 1 may be simplified. In this case, the face image data of the user captured by each portrait creating apparatus 1 is transmitted to the server 32, and the face part image data selected by the server 32 is transmitted to each portrait creating apparatus 1.

Further, in the databases 53 to 55, measured values extracted from the respective face parts are stored as feature amounts instead of the face part codes, and retrieval is performed based on the measured values of the respective face parts extracted from the user's face image. , A more detailed search result may be obtained.

In addition, in the above embodiment, the search content is the same or similar to a beautiful face, a celebrity, or another registered user. However, the database 41 of the server 32 is provided with a database corresponding to the search content. Alternatively, a search process can be performed for other search contents.

The mode of using the search result in the portrait creating apparatus 1 is not limited to the display on the display. For example, a portrait image in which an image corresponding to the search result is synthesized is created, and the created result is displayed. You may make it output according to a user's selection.

It should be noted that the present invention is not applied only to face image retrieval, but is applied to crops and industrial products.
Alternatively, it can be used for a search process for an article from which two or more feature values can be extracted. For example, when searching for a specific type of crop from among the harvested crops using the overall size and color etc. as features, the overall size, color, size of eyes and eye position etc. are landed as features. The present invention is also applicable to a case of searching for a specific type of fish from among a large number of types of fish that have been performed, and a case of searching for a specific type of vehicle from a running vehicle using the axial distance, the total length, the total height, and the like as features. Can be applied.

[0052]

According to the first, eighth, and fifteenth aspects of the present invention, an image having the same or similar feature as an image to be searched is searched from a plurality of images stored in a feature storage unit. By doing so, it is possible to objectively and quantitatively determine the identity and similarity of the images based on the feature amounts of the image of the target object such as dimensions and angles.

According to the second, ninth and sixteenth aspects of the present invention, a feature amount is extracted from the input image of the object, and an image having the same or similar feature amount as the extracted feature amount is stored in the feature amount. By performing a search in the section, it is possible to objectively and quantitatively determine the identity and similarity between the input image and a known image stored in advance in the feature amount storage section.

According to the third, tenth, and seventeenth aspects, the feature amount extracted from the input image of the object is stored in the feature amount storage unit, and is the same as or similar to the preset reference image. By retrieving images in the feature storage unit, it is possible to objectively and quantitatively determine the identity and similarity of the input plurality of target images with the reference image.

According to the fourth, eleventh and eighteenth aspects of the present invention, by receiving input of a large number of objects from a plurality of terminal devices, a search for a known image or an image identical to or similar to a reference image is performed. The hit rate in the above can be improved.

According to the fifth, twelfth, and nineteenth aspects of the present invention, the same or similar image is determined based on a combination of characteristic amounts such as dimensions and angles individually extracted for each component of the image of the object. By making the determination, the identity and similarity of the images can be determined as the sum of the determination results of the identity and similarity of each of the constituent parts in the image of the object, and the accuracy of the determination results is improved. Can be.

According to the sixth, thirteenth, and twentieth aspects of the present invention, image component data corresponding to dimensions, angles, and the like individually extracted for each of the constituent parts of the image of the object are selected, and a plurality of image components are selected. By judging the identity and similarity of images using a set of data as a feature amount, it is possible to simplify the determination of the identity and similarity of images based on the classified image component data of each component. .

According to the seventh, fourteenth, and twenty-first aspects of the present invention, face part data of a portrait corresponding to dimensions, angles, and the like individually extracted for face constituent parts such as contours, eyes, nose, and mouth. And determining the sameness and similarity of the face images based on the portrait image which is a set of a plurality of pieces of face part data, thereby improving the entertainment in the determination of the sameness and similarity of the face images. it can.

[Brief description of the drawings]

FIG. 1 is an external view of a caricature creation apparatus to which a face image search system, a face image search method, and a face image search program storage medium according to an embodiment of the present invention are applied.

FIG. 2 is a block diagram showing a configuration of the portrait drawing apparatus.

FIG. 3 is a diagram showing storage contents of a database of the portrait creating apparatus.

FIG. 4 is a diagram showing a configuration of a face image search system including the portrait creating apparatus.

FIG. 5 is a diagram showing storage contents of a database of registered users possessed by a server constituting the face image search system.

FIG. 6 is a diagram showing storage contents of a first reference face image database of the server.

FIG. 7 is a diagram showing storage contents of a second reference face image database of the server.

FIG. 8 is a diagram showing storage contents of a same face image database of the same server.

FIG. 9 is a flowchart showing a processing procedure of a control unit of the portrait creating apparatus.

FIG. 10 is a flowchart showing a processing procedure of a control unit of the server.

[Explanation of symbols]

 1-caricature creation device 11-control unit 14-imaging unit 16-face part database 31-face image search system 32-server 42-control unit 42a-registration engine 42b-search engine 52-registered user database 53-first reference Face image database 54-Second reference face image database 55-Same face image database

Claims (21)

[Claims] 1. A feature amount storage unit that stores feature amounts of images of a plurality of objects having substantially similar shapes, wherein the feature amount is the same as or similar to the search target image based on the feature amount of the search target image. An image search system, comprising: a feature amount search unit that searches an image that can be determined in a feature amount storage unit. 2. An image input unit for receiving an input of an image of an object, and a characteristic amount extracting unit for extracting a characteristic amount from the image of the object input from the image input unit, wherein the characteristic amount searching unit is The image search system according to claim 1, wherein the feature amount storage unit searches for images having the same or similar feature amounts extracted by the feature amount extraction unit. 3. An image input unit for receiving an input of an image of an object, and a characteristic amount extracting unit for extracting a characteristic amount from an image of the object input from the image input unit, wherein the characteristic amount storage unit is provided. The feature amount of each target object extracted by the feature amount extraction unit is stored, and the feature amount search unit stores an image having the same or similar feature amount as the feature amount of a preset reference image. The image search system according to claim 1, wherein the search is performed by a section. 4. A server provided with said feature amount storage unit and said feature amount search unit is connected to a plurality of terminal devices each provided with said image input unit and said feature amount extraction unit. The image search system according to claim 2, wherein the feature amount data and the search data of the image of the target object are transmitted and received between the images. 5. The image retrieval system according to claim 1, wherein the feature amount is a combination of feature amounts individually extracted for each component of the image of the object. 6. A component storage unit in which a plurality of image component data classified in accordance with a feature amount for each component of an image of an object is stored in advance, and the feature amount is a component of the image of the object. The image retrieval system according to any one of claims 1 to 5, wherein the image retrieval data is image component data selected from the component storage unit based on the characteristic amount of the image component. 7. The image retrieval system according to claim 6, wherein said part storage unit stores a plurality of face part data classified according to the feature amount of each of the constituent parts of the face image as portrait image data. 8. A feature amount storage unit for storing feature amounts of images of a plurality of objects having substantially similar shapes, and determining that the images are the same or similar to the search target image based on the feature amounts of the search target image. An image search method, wherein an image that can be determined is searched in a feature amount storage unit. 9. An image receiving apparatus receives an input of an image of an object, extracts a feature amount from the input image of the object, and stores images having the same or similar feature amounts in the extracted image of the object. The image search method according to claim 8, wherein the search is performed in: 10. An input of an image of an object image is received,
The feature amount is extracted from the input image of the target object, and the feature amount of the extracted image of the target object is stored in the feature amount storage unit, and the same or similar feature amount as the preset feature amount of the reference image is stored. 9. The image search method according to claim 8, wherein the image is searched for in the feature amount storage unit. 11. A server for storing a feature amount of an image of an object and retrieving an image having a feature amount that is the same as or similar to a feature amount of a preset reference image; The image search method according to claim 9, wherein the method further comprises transmitting and receiving feature amount data and search data to and from a plurality of terminal devices that receive the input and extract a feature amount from the input image of the target object. 12. The image retrieval method according to claim 8, wherein said characteristic amount is a combination of characteristic amounts individually extracted for each component of the image of the object. 13. A plurality of image component data classified in accordance with a feature amount for each component of an image of a target object in a component storage unit in advance, and a plurality of image component data corresponding to the feature amounts of each component portion of the image of the target object are stored. The image retrieval method according to claim 8, wherein the image component data to be selected is selected from the component storage unit as the feature amount. 14. A caricature image data is stored in the part storage unit as a plurality of face part data classified for each of the constituent parts of the face according to the feature amount.
3. The image search method according to 3. 15. A process for storing feature amounts of images of a plurality of objects having substantially similar shapes in a feature amount storage unit, wherein the process is performed on the basis of the feature amounts of the search target images in the same or similar manner as the search target images. An image search program recording medium characterized by recording a program comprising: a process of searching an image which can be determined to be present in a feature amount storage unit. 16. A process for receiving an input of an image of an object, a process for extracting a feature amount from the input image of the object, and storing an image having the same or similar feature amount as the extracted feature amount. 16. The image search program recording medium according to claim 15, wherein a program including a process of searching in a unit is recorded. 17. A process for receiving an input of an image of an object, a process for extracting a feature amount from the input image of the object, and a process for storing the feature amount of each extracted image in the feature amount storage unit. 16. The image search program storage medium according to claim 15, wherein a program including a process of searching the feature amount storage unit for an image having the same or similar feature amount as a preset reference image feature amount is recorded. 18. A server for storing a feature amount of an image of an object and retrieving an image having the same or similar feature amount as that of a preset reference image; 18. The program according to claim 16, wherein a program including a process of transmitting and receiving feature amount data and search data between a plurality of terminal devices that accepts and extracts a feature amount from an input image of the target object is recorded. Image search program recording medium. 19. The image search program recording medium according to claim 15, wherein said feature amount is a combination of feature amounts individually extracted for each component of the image of the object. 20. A program which includes a process of selecting image component data, which is image component data stored in advance in a component storage unit and corresponding to a feature value of each component of an image of a target object, as the feature value. An image search program recording medium according to any one of claims 15 to 19. 21. The image search program recording medium according to claim 20, wherein said image part data is portrait image data, and a program including a process of selecting portrait image data as said feature amount is recorded.