JPH10340272A - Simular picture retrieval device/method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make it possible efficiently narrow down candidate pictures containing a target picture by indicating plural simular pictures or related pictures. SOLUTION: For accumulating the pictures inputted by a picture input part 1 in a picture accumulation part 2, it is registered in a cache memory 42 in accordance with the designation of a user. The list of the pictures registered in the cache memory 42 is displayed on a picture display part 3 by a picture indication part 4 and the picture designated by the user is selected from it. When the designation of a retrieval element area from the user making the selected picture to be an object is received, information of the designated area is added to the pertinent picture in the cache memory 42. The picture indication part 4 selects the plural pictures whose retrieval element areas are previously set and which are designated by the user are selected for defining the picture retrieval condition and a picture retrieval part 5 retrieves the simular picture from the picture accumulation part 2 based on the respective retrieval element areas of the plural selected pictures.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a similar image retrieval apparatus and method suitable for efficiently retrieving a target image from a large number of images.

[0002]

2. Description of the Related Art In recent years, multimedia information-related computer systems have been used on an individual basis. In addition to conventional text, media such as voice, graphics, color images, and video can be easily used as media. It can be handled. Such image materials are available on the Internet World Wide Web (WW).
W) There are many free sites on each site above and on paid image databases. From this large amount of image materials (image data), the work of searching for an image that each individual needs, for example, an image that suits each individual's preference and interval is extremely burdensome.

Conventionally, in order to facilitate an image search operation, a plurality of reduced search target images are displayed on one screen at the same time to allow a user to select the image, or a feature of an image to be searched by the user. Has been applied as a search condition that allows a user to present a sketch image representing a target image as a search condition, or one image that is similar to a target image as a search condition by a user operation.

[0004]

However, in a conventional search method, for example, a method using so-called multi-image display, in which a plurality of reduced images are displayed on a screen at the same time and a user selects one of the reduced images is selected. If the number of pieces of image data exceeds a certain level, the number of display screens increases, and the search efficiency drops immediately. Furthermore, when the number of reduced screens increases, there is also a problem that the search must be performed while switching the reduced screens and watching for a long time.

[0005] Next, in the method of presenting similar images using simple sketch images, it is difficult to provide accurate images unless there is a certain sense of drawing, and drawing such a pattern is troublesome even if it is a sketch image. There is such a problem.

On the other hand, an image similar to a target image is
The method of presenting a sheet is a simpler method than the above two methods. By the way, searching for a similar image does not mean searching for a specific image that is clear, but, for example, in order for the user to select a desired image material, an image according to the user's own sensory taste is selected. There is a case where it is desired to present the image as it is, search for an image group similar to the image, and obtain an image that more sensuously matches from the similar image group. Even in such a case, it is convenient if the method of presenting one image can be applied. However, in the search operation by presenting one image, it is possible to search a group of similar images to some extent, but for example, an image with a green grove on the left side of the image and a bright red sunset on the upper right side. As described above, when an attempt is made to search using a plurality of search elements, a complicated operation is performed in which the operation is performed in multiple stages while sequentially switching the specified search element.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and an object of the present invention is to present a similar image or related images as a plurality of images in a system handling a large amount of image materials. Another object of the present invention is to provide a similar image search apparatus and method capable of efficiently narrowing down candidate images including an image to be searched.

[0008]

According to the present invention, there is provided an image storage means for storing an image, an image specified by a user, and an area or an object serving as an image search key in the selected image. Search element setting means for setting as a search element area or a search element according to the user's specification, and user's specification of a plurality of images specified by the user in which the search element area or the search element is set by the search element setting means Search condition setting means for setting an image search condition definition in accordance with the above, and the image storage based on each search element region or search element of a plurality of images set for the image search condition definition by the search condition setting means. Image search means for searching for a similar image from the means.

In such a configuration, for each of the plurality of images specified by the user, a search element area or a search element is set according to the user's specification, and the search element area or the search element is set. By using multiple images specified by the user for defining image search conditions according to the user specification, it is possible to set image search conditions that match the user's sense, Can be efficiently searched by simple operation.

In particular, an image input means for inputting an image is provided, and the image input by the image input means is stored in the image storage means, and the input image or the image stored in the image storage means is stored. A search key candidate image registration unit for registering a plurality of images that can be a key for image search in accordance with a user's designation, and a plurality of images registered here are displayed by an image list presentation unit. If the configuration is such that the user can select and specify an image for setting a search element area or a search element to be used as an image search key and present the image to the user, the user can use the image to be used as the image search key. It can be made easily selectable.

Also, the setting of the search element area or the search element and the setting for defining the image search condition are separated, and the search element area or the search element is set in advance regardless of the setting for defining the image search condition. It is good also as a structure which can be stored. In this case, as the data registration structure of the image in the search key candidate image registration means, whether or not the search element area or the search element has been set, and if set, the set search element area or the search element can be identified. Apply the structure. Also, when presenting the image list by the image presenting means, whether or not the search element area or the search element has been set, and if so, the set search element area or the search element itself can be visually recognized by the user. In this way, it may be expressed by highlight display or specific color display.

[0012] This makes it possible to set a search element area or a search element at an arbitrary time for an image registered in the search key candidate image registration means.
In addition, the user can visually recognize whether or not the search element area or the search element has been set from the presented image list, and if so, the set search element area or search element can be visually recognized. Can easily be specified.

The image search means is connected to each of the search element areas or search elements of the plurality of images set for defining the image search conditions by the search condition setting means by a logical operator designated by the user. It is preferable to have a function of generating a search condition expression composed of the following elements, and to search for similar images based on the search condition expression. In particular, a function for displaying a plurality of logical operation boxes of different logical operation types on the screen for setting a user-specified image to be subjected to a predetermined logical operation on the search element area or the search element is provided by the search condition. In the image search means, the search element area or the search element of the user-specified image set in each of the above-mentioned logical operation boxes is subjected to a logical operation unique to the (user-selected) box. That is, by performing a logical operation selected by the user) and generating a search condition expression composed of a plurality of elements based on the result of each logical operation, a search condition setting operation using a plurality of images by the user is performed. Can be simplified, and an excellent user interface can be realized.

[0014] Further, a feature of each search element area or a search element of a plurality of images set for defining an image search condition by the search condition setting means is extracted as color information, and the color information is used for similar image search by the image search means. The image search condition can be simplified by using the image search condition.
Here, if the color information is decomposed into at least components of hue, saturation, and lightness, and a predetermined component or a user-specified component among them is used as an image search condition, the image search process can be simplified. It is possible to set a matching image search condition according to the user's sense while trying to achieve the image conversion. At this time, each image is divided into n × m blocks, and representative color information (for example, the average value, intermediate value, Or maximum)
The similarity of the search element region or the entire search element may be obtained using the representative color information in block units.
Also, for all the images stored in the image storage means, representative color information of all n × m blocks is extracted in advance,
It is stored in association with the image, and when searching,
It is also possible to adopt a configuration in which the representative color information of each block corresponding to the search element area or the search element is extracted and used.

[0015]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of a similar image search device according to one embodiment of the present invention.

In FIG. 1, an image input unit 1 takes in an image via the Internet or an image from another image database according to a user's operation. The image captured by the image input unit 1 is stored in the image storage unit 2 via an image search unit 5 described later, for example. The image may be captured in a batch or in the Internet while surfing the Internet.

The image captured by the image input unit 1 is displayed on the image display unit 3 via the image presentation unit 4. The user selects a characteristic image, an image that remains impressive, or an image including an object to be used from among the captured images displayed on the image display unit 3 by operating a pointing device such as a mouse. Can be specified.

The image presentation unit 4 has a user interface 41 for interacting with a user via the image display unit 3 and a cache memory 42 for storing images usable as search conditions for similar images. ing. Image presentation unit 4
Detects a user's image selection operation through the user interface 41, and registers the image specified by the selection operation in the cache memory. Here, if the cache memory 42 is full (FULL), an image that has never been used is deleted from the registered images, and
An image specified by the user is newly registered in the area. If all the images have been used once, the image used most recently is deleted and an image specified by the user is newly registered in that area. In addition, the user may be allowed to select an image to be deleted. In addition, the deletion of the registered image from the cache memory 42 can be performed independently of the new registration of the image. The images already stored in the image storage unit 2 may be presented to the user through the image display unit 3, and the image selected by the user from the images may be registered in the cache memory 42.

The image presenting unit 4 presents the image registered in the cache memory 42 to the user, and interacts with the user via the image display unit 3 using the user interface 41, thereby providing the user with the image. A similar image search condition specified by a user for searching for an intended similar image is received. For this purpose, the image presenting unit 4 includes an image list presenting function of presenting the images registered in the cache memory 42 to the user, and an image element that can be used as a search key with a user-specified image element targeting the presented image ( Search element setting function to accept (set) as a search element) and an image with a user-specified search element required to generate a similar image search formula (search condition) for searching for similar images intended by the user It has a search condition setting (definition) function capable of accepting a logical operation type for presenting a plurality of images and connecting between the presented images (search elements in the images).

The image search unit 5 performs similar image search based on an image having a user-specified search element presented through the search condition setting function of the image presenting unit 4 and a logical operation type specified by the user. For example, a search formula is generated, a large amount of images stored in the image storage unit 2 are searched according to the search formula, and the result is displayed on the image display unit 3.

The system control unit 6 includes the image input unit 1,
It controls the components of the image storage unit 2, the image display unit 3, the image presentation unit 4, and the image search unit 5, and manages and controls the components. Next, the operation of the configuration of FIG. 1 will be described with reference to FIGS.

First, it is assumed that an operation related to similar image retrieval is requested to the apparatus of FIG. 1 by a user operation. In this case, the system control unit 6 activates the image presentation unit 4. Then, the image presenting section 4 (the user interface 41 therein) displays on the image displaying section 3 an image presenting / retrieving image list presenting area 21, a search element setting area 22, and a search condition setting area 23 as shown in FIG. Condition setting screen 2
0 is displayed (step S1 in FIG. 5). Here, in the image list presentation area 21, for example, a group of images registered in the cache memory 42 is displayed in a reduced size. If the number of images registered in the cache memory 42 is large and cannot be displayed in the image list presentation area 21 at one time, the images are displayed by scrolling or by switching screens. For example, a search element setting area 22 is provided on the left side of the image list presentation area 21.

The user, when there is an undefined image to be defined as an image element that can be used for search in the image group displayed in the image list presentation area 21, sets the image to the search element setting area 22. Perform an operation to browse up to.
As a result, the image presenting unit 4 displays the user-specified image on the image list presenting area 21 in the search element setting area 22, and receives designation of a target area (search element area) for the specified image (FIG. 5). Steps S2 to S4). A “setting” button 221 and an “all” button 222 are arranged, for example, below the search element setting area 22.

When the user sets the entire image browsed to the search element setting area 22 as the search element area, the user clicks the “whole” button 222 with a mouse or a pen, and then sets the “set” button 221. Click.
Also, the desired object (the area where the object exists) in the image
Is set so as to surround the object with a mouse, a pen, or the like, and a “setting” button 221 is clicked.

Upon detecting that the “SET” button 221 has been clicked, the image presenting section 4 (the user interface 41 therein) replaces the image on the search element setting area 22 with the position on the original image list presenting area 21. Return to At this time, when the “whole” button 222 is clicked, that is, when the entire image is designated as the search element area, the image presenting unit 4 marks the image frame with a predetermined color, for example, red, and If the area where the object exists is specified as the search element area, the area specified by the user,
That is, the locus surrounded by a mouse, a pen, or the like is marked with a predetermined color (here, red as in the case of the entire designation) (step S5 in FIG. 5).

In this way, the user can select an image defined as a search element usable for search and an undefined image from among the images displayed in the image list presentation area 21. Easy to see. Moreover, regarding the definition image, it is possible to visually recognize which of the whole or the part (object) is set as the search element area, and in the case of the part (object), which area on the image is set. .

The above-described definition image, undefined image, and the search element area in the definition image must be identifiable in the cache memory 42. Then, the image setting unit 4 sets (defines) the entire image or the user-specified area browsed to the area 22 when the “setting” button 221 on the search element setting area 22 is clicked as the search element area. Is instructed, the corresponding image (data) in the cache memory 42 has
Flag information (definition / undefined identification flag) 41 indicating that the image is a definition image, and flag information (search element range identification flag) indicating whether the search element area is the entire image or a part (object) ) 42 and area information 43 indicating the search element area when the area is a part (step S6 in FIG. 5). The addition here includes a logical addition, and may be linked to the corresponding image data by a pointer or the like.

In the present embodiment, the area information 43 of the search element area does not indicate the area designated by the user, that is, the area itself surrounded by a mouse, a pen, or the like.
It is generated by the image presentation unit 4 as follows. That is, in the present embodiment, the entire image in which the search element area is defined is divided into n × n (n is an integer of 1 or more) blocks, and a user partially or partially divides the area into an area surrounded by a mouse, a pen, or the like. A list of the positions (block numbers representing the blocks) including all the blocks is used as the area information of the search element area. The search element area in the following description may refer to an area represented by the list of block positions, or may refer to an area itself surrounded by a mouse or a pen by a user.

In this embodiment, search conditions for similar images based on a plurality of presentation images can be set on the search condition setting area 23. The images available in the search condition setting area 23 are the image list presentation area 21
Only the definition image of the above presented image group, that is, the image set in the search element setting area 22. That is, in the present embodiment, although the definition image in the group of presentation images on the image list presentation area 21 can be browsed to the designated position in the search condition setting area 23 according to the user's selection / designation operation (FIG. 6).
Steps S11 to S13), the undefined presentation image is not browsed in the search condition setting area 23.

In the search condition setting area 23, a format for a search formula (search logical formula) is determined in advance. In this example, as shown in FIG. 2, a logical product box 231 and a logical sum box 232 are prepared, and an “OR” button 233 and an “AND” button 234 are arranged between them. Both boxes 231, 232 are enclosed in parentheses,
An “OR” button 233 or an “AND” is used to define a partial search expression according to the setting in the logical product box 231 and a partial search formula according to the setting in the logical sum box 232.
A search expression linked by a logical operator indicated by the button that is clicked among the buttons 234 can be set. Here, "OR" button 233 or "AND"
When the button 234 is clicked, a predetermined display modification is applied to the button, and it is specified that the button is clicked (steps S14 and S15 in FIG. 6).

In the search condition setting area 23, a "search" button 235 is arranged. In this embodiment,
By clicking the “search” button 235 (step S16 in FIG. 6), the operation of generating a search expression based on the settings in the search condition setting area 23 (step S18 in FIG. 6) and the generation of a similar image based on the generated search expression The search operation (step S19 in FIG. 6) is started.

First, a description will be given of a search expression generating operation using the search condition setting area 23. Image presentation unit 4
Is a "search" button 235 in the search condition setting area 23.
Is clicked, a predetermined display modification is applied to the "search" button 235 to indicate that the button 235 has been clicked (step S17 in FIG. 6), and control is passed to the image search unit 5. At this time, the search condition setting area 23
It is assumed that two definition images of an image A and an image B are browsed from the image list presentation area 21 in FIG. 2 in the AND box 231 of FIG. In the image A, a lower left area, for example, a green grove in the image A is set as the search element area 31, and in the image B, an upper right area, for example, an orange sunset area in the image B is the search element area 32. Is set as

On the other hand, in the OR box 232 of the search condition setting area 23, it is assumed that only the image C is browsed from the image list presentation area 21 in FIG. 2 as shown in FIG. This image C is, for example, an image in which a red motor boat is floating on a lake, and the area of the red motor boat is set as the search element area 33.

In this embodiment, in order to simplify the search processing in the image search unit 5, each search element area is handled by the color information of the image (instead of the shape of the object in the area). Therefore, in the search expression generation processing in step S18, the image search unit 5 sets the search condition setting area 2
For each of the images browsed in the logical product box 231 and the logical sum box 232 of 3, the representative color information of the block is determined for each block constituting the set search element area (step S21 in FIG. 7). . That is, the image search unit 5 converts each pixel data (RGB data) of all the pixels in the target block into three color attributes (components) of, for example, hue (H), saturation (S), and lightness (I). By decomposing, a predetermined value of an average value, an intermediate value (an intermediate value between the minimum value and the maximum value), or a maximum value is calculated for each attribute, and the value calculated for each attribute is calculated for the corresponding block. Is the representative value of the attribute. That is, the image search unit 5 obtains a representative value for each attribute of the hue (H), the saturation (S), and the brightness (I) in units of each block constituting the set search element area. Therefore, in the example of FIG. 3, the hue (H) is set for each block forming the search element area 31 of the image A, each block forming the search element area 32 of the image B, and each block forming the search element area 33 of the image C. ), Saturation (S),
A representative value is extracted for each attribute of lightness (I).

When the "search" button 235 is clicked, the image search unit 5 sets the hue (unit) for each block constituting the search element areas 31 and 32 of the images A and B set in the AND box 231. H), saturation (S), and brightness (I) are extracted for each attribute (step S21 in FIG. 7), and the images A and B having the extracted attribute value blocks are assigned “AN”.
By applying the “D” process, a partial search formula according to the setting content of the logical product box 231 is generated (Step S22 in FIG. 7). Here, for the sake of simplicity, it is assumed that a representative color is used instead of the representative value of the attribute, and that all blocks forming the search element areas 31 and 32 are green and orange, respectively. The search element area 3 corresponding to a certain search element area 31 is green and the sunset of the image B is present.
A partial search expression indicating that an image corresponding to area 2 is searched for an orange image is generated.

Further, the image search unit 5 is provided with a logical sum box 23.
Hue (H), saturation (S), lightness (I) in units of each block forming the search element area 33 of the image C set to 2
A representative value is extracted for each attribute (step S21 in FIG. 7), and an “OR” process is applied to the image C having the extracted attribute value block, thereby obtaining a partial search expression according to the setting contents of the logical sum box 232. Is generated (step S2 in FIG. 7).
2). Here, for the sake of simplicity, it is assumed that a representative color is used instead of the representative value of the attribute, and that all the blocks (representative color) forming the search element area 33 are red, there is a motor boat of pixel C. A partial search expression indicating that an area corresponding to the search element area 33 searches for a red image is generated.

The image search unit 5 includes boxes 231 and 232
Is generated, a partial search expression is generated according to the setting contents of “O” in the box 231 or 232 between the partial search expressions.
By applying the logical operation process indicated by the button that has been clicked among the "R" button 233 or the "AND" button 234, a search formula according to the setting contents of the search condition setting area 23 is generated (step S23 in FIG. 7). In the example of FIG. 3, the “OR” button 233 is in a state where a predetermined display modification is performed. This indicates that the “OR” button 233 has been clicked.

In this case, the search expression generated by the image search unit 5 is “AND” indicated by a logical product box 231.
A search condition for searching for an image that satisfies the condition or an image that satisfies the “OR” condition indicated by the OR box 232 is shown. Therefore, when the example of the simplified description is applied here, the area corresponding to the search element area 31 with the prominence of the image A is green and the search element area 32 where the sunset of the image B is present. A search formula is generated indicating that an image whose area is an orange image or an image corresponding to the search element area 33 with the motorboat of the image C is searched for a red image. That is, in the example of the search condition setting area 23 in FIG. 3, an image in which two areas having similar positions and colors exist in the green search element area 31 and the orange search element area 32, respectively, or the red search element area 33 Means that a search condition for searching for an image in which an area having a similar position and color exists is set.

On the other hand, if the two buttons 233 and 234 between the boxes 231 and 232 in FIG.
If the “AND” button 234 is clicked, an area having a similar position and color to the green search element area 31, an area having a similar position and color to the orange search element area 32, and a red Means that the search condition for searching for an image in which all the regions having similar positions and colors are present in the search element region 33 is set.

The image search section 5 includes a search condition setting area 23
When a search formula is generated in accordance with the setting of, image search processing (step S19 in FIG. 6) for searching the image storage unit 2 for an image that matches the search condition indicated by the search formula is performed as follows.

First, the image search unit 5 selects one image from the image storage unit 2 as a search target image, and
The block is divided into × n blocks, and all blocks in the mask area are cut out from the search element area of each image satisfying the search condition (set through the search condition setting area 23) as a mask area. Next, the image search unit 5 sets the average value, the intermediate value, and the representative color information (here, representative values of the three color attributes H, S, and I) of the extracted block as a unit. Alternatively, a predetermined value among the maximum values is obtained. Next, the image search unit 5 obtains color information obtained for each block in the search element area of each image forming the search condition as a unit, and uses the search element area as a mask area for each block from the search target image. From the obtained color information, the similarity for each area is calculated. Next, the similarity of the entire image is obtained by applying the similarity for each region to the previously generated search formula.

At this time, instead of using H, S, and I fixedly, an arbitrary combination may be selected and designated by the user. Further, it is preferable that the similarity for each of the color attributes of H, S, and I is weighted so that each weight can be specified by a user. In addition, the pixel data (RGB data) may be decomposed into four color attributes of hue (H), saturation (S), lightness (I), and complexity (C).

The image search section 5 repeats the above processing for all the images stored in the image storage section 2 (however, the image serving as a search key may be excluded).
An image having a similarity greater than or equal to a predetermined threshold is selected as a similar image and displayed on the image display unit 3.

As described above, the image storage unit 2 uses the condition based on the search formula generated according to the setting of the search condition setting area 23.
By performing an image search for, similar image groups that match the sense of the image intended by the user can be efficiently narrowed down and extracted as candidates, and presented to the user through the image display unit 3. The group of candidate images presented to the user resembles the sensation of the image initially aimed at, and from this, the image that best matches the sensation can be finally selected. This is an effect in consideration of ambiguity in the image search applied in the present embodiment.

In the above embodiment, the image search unit 5
Has been described as obtaining representative color information in units of blocks in the mask area each time an image to be subjected to an image search is retrieved from the image storage unit 2. However, the present invention is not limited to this. For example, for all the images stored in the image storage unit 2, the entire image is divided in advance into n × n blocks, representative color information is obtained in block units, and the image storage unit 2 may be stored. In this case, at the time of the image search by the image search unit 5, the representative color information of each block in the mask area of the search target image can be obtained at an extremely high speed.

The value of n may be fixed or variable. For example, if the value of n can be set arbitrarily according to the user's specification, the value of n may be increased to increase the shape. The processing speed can be further increased by increasing the precision of the expression or reducing the value of n. Also,
Instead of n × n, n × m (where n and m are different 1
It may be divided into blocks of the above (integer).

In the above embodiment, the case where the user directly specifies the search element area (area) on the search element setting area 22 has been described. However, an object to be used as a search element is represented (for example, a motor boat). Word information may be input from a keyboard or the like or by voice. However, in this case, it is necessary to provide a shape recognition unit that recognizes the object indicated by the input word information from the image in the search element setting area 22 and sets the area of the recognized object as the search element area. As described above, when the method of recognizing the shape of the object specified by the user is applied, the search element area determined from the shape does not need to be a fixed position in the image, and can be dealt with by a relative positional relationship. Thus, it is possible to perform the search processing more efficiently. In this case, it is only necessary to search for images having similar shapes only, and it is not always necessary to find the representative colors in block units.

In the above embodiment, a case has been described where a plurality of images are presented in the search condition setting area 23 according to a user's instruction. However, one image is stored in either the logical product box 231 or the logical sum box 232. It is also possible to cope with a case where only one is presented.

In the above embodiment, the case where the user specifies one search element (including a search element area) for each image presented in the search condition setting area 23 has been described. It is also possible to set search elements. For example, if a green tree area and an orange sunset area exist in one presentation image, both areas can be registered as search elements.

In the above embodiment, a case has been described in which an object in an image, that is, image information itself, is used as a search element and an image search is performed by comparing image information (here, comparing color information). A more efficient search can be performed by a combination with another search means or a search element depending on the search purpose. For example, since a human can easily recall a memory by organizing information in a time series, the image storage unit 2 is used as a search element.
By using the date range by using calendar information such as the date of image registration to the user, the number of search candidates can be narrowed down appropriately, and image candidates that match the sense can be searched.

Further, in the above embodiment, the search condition setting area 23 used for setting the search condition has been described assuming that the format for the search formula (search logical formula) is predetermined. It is more convenient to be able to change the format for image search depending on the purpose of image search.Therefore, preparing multiple types of search expression definition format patterns in the form of templates and allowing the user to select from them Good.

The processing procedure in the similar image search apparatus applied in the above embodiment is a CD-RO storing a program for causing a computer such as a personal computer capable of reading the program to execute the processing procedure.
This is realized by causing a computer to read and execute a program recorded on a recording medium such as an M, a DVD-ROM, a floppy disk, and a memory card.
The content of the recording medium on which the program is recorded may be downloaded to a computer via a communication line or the like.

[0053]

As described above in detail, according to the present invention, a plurality of images specified by a user can be displayed by an interactive user interface with the user.
Set a search element area or search element according to the user's specification, and use multiple images specified by the user with the search element area or search element set for defining image search conditions according to the user's specification Therefore, it is possible to set an image search condition that matches with the user's sense, so that a search with a plurality of search elements intended by the user can be efficiently performed by a simple operation.

[Brief description of the drawings]

FIG. 1 is a block diagram showing the overall configuration of a similar image search device according to an embodiment of the present invention.

FIG. 2 is an exemplary view showing an example of an image presentation / search condition setting screen applied in the embodiment.

FIG. 3 is a view showing a search condition setting example using a search condition setting area in FIG. 2;

FIG. 4 is a diagram showing an example of information added to a corresponding image when a search element area is set.

FIG. 5 is a view showing a part of a flowchart for explaining the operation of the embodiment;

FIG. 6 is a view showing the rest of the flowchart for explaining the operation of the embodiment.

FIG. 7 is a flowchart showing a detailed procedure of a search expression generation process in step S18 in FIG. 6;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Image input part 2 ... Image storage part 3 ... Image display part 4 ... Image presentation part (image list presentation means, search element setting means, search condition setting means) 5 ... Image search part 6 ... System control part 20 ... Image presentation -Search condition setting screen 21 ... Image list presentation area 22 ... Search element setting area 23 ... Search condition setting area 41 ... User interface 42 ... Cache memory (search key candidate image registration means) 231 ... Logical product box 232 ... Logical sum box

Claims (9)

[Claims] An image storage means for storing an image, an image specified by a user is selected, and an area or an object serving as an image search key is searched in the selected image in accordance with the specification of the user. A search element setting means for setting as an element area or a search element; and a plurality of user-specified images in which the search element area or the search element is set by the search element setting means for defining an image search condition according to the user's specification. Search condition setting means for setting, and an image for searching for a similar image from the image storage means based on each search element area or search element of a plurality of images set for image search condition definition by the search condition setting means A similar image search device comprising a search unit. 2. An image input device for inputting an image, an image storage device for storing an image input by the image input device, and an image input by the image input device or the image storage device. A search key candidate image registration unit for registering a plurality of images that can be used as image search keys from among the stored images in accordance with a user's designation; and a plurality of images registered in the search key candidate image registration unit. An image list presenting means for presenting a plurality of images to the user, selecting an image specified by the user from the images presented by the image list presenting means, and an area serving as an image search key in the selected image or Search element setting means for setting an object as a search element area or a search element in accordance with a user's designation; A search condition setting unit configured to set a plurality of images specified by a user for image search condition definition in accordance with a user's specification among the images in which the elements are set; and for the image search condition definition by the search condition setting unit. A similar image search device comprising: an image search unit that searches for a similar image from the image storage unit based on each search element area or search element of a plurality of set images. 3. The image search means connects each search element region or search element of a plurality of images set for image search condition definition by the search condition setting means with a logical operator specified by a user. 3. The similar image search device according to claim 1, wherein a search condition expression including the following elements is generated, and a similar image is searched from the image storage unit based on the search condition expression. 4. A method according to claim 1, wherein said search condition setting means sets a plurality of logics of different logical operation types for setting a user-specified image to be subjected to a predetermined logical operation on the search element area or the search element. Displaying a calculation box on the screen, the image search means performs a unique logical operation on the search element region or search element of the user-specified image set in each of the logical operation boxes, 3. The similar image search device according to claim 1, wherein a search condition expression including a plurality of elements is generated based on a result of the logical operation. 5. The image search means extracts, as color information, a search element area or a feature of each search element of a plurality of images set for image search condition definition by the search condition setting means, and extracts the color information. 5. The similar image search device according to claim 1, wherein the similar image search device is used for an image search condition. 6. The image search means decomposes the color information into at least components of hue, saturation, and brightness, and uses a predetermined component or a user-specified component among the components for an image search condition. 6. The similar image search device according to claim 5, wherein: 7. The search element setting unit according to claim 2, wherein the search element setting unit adds information of the set search element area or search element to a corresponding image in the search key candidate image registration unit. Similar image search device. 8. A similar image search method for searching for an image similar to an image intended by a user from a group of images stored in an image storage means, wherein an image specified by the user is selected and selected. Setting an area or an object serving as an image search key in the obtained image as a search element area or a search element according to a user's specification; and a plurality of user-specified images in which the search element area or the search element is set. Setting for an image search condition definition according to a user's designation; and a similar image from the image storage unit based on each search element area or search element of a plurality of images set for the image search condition definition. Searching for similar images. 9. A computer-readable recording medium which records a program for searching for an image similar to an image intended by a user from a group of images stored in an image storage means, wherein Selecting the selected image, setting an area or an object serving as an image search key in the selected image as a search element area or a search element in accordance with a user's specification; and setting the search element area or the search element. Setting a plurality of images specified by the user for image search condition definition according to the user's specification; and setting each of the search element areas or search elements of the plurality of images set for the image search condition definition. And a computer-readable storage medium storing a program for causing a computer to execute a step of searching for a similar image from the image storage means. .