JP2006004157A - Image retrieval program, image retrieval method, image retrieval device, and storage medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image retrieval program, an image retrieval method, an image retrieval device and a storage medium capable of extracting a desired image as a retrieval result, including an image with which a keyword or not is matched. <P>SOLUTION: This image retrieval method comprises selecting a plurality of image files stored in an image storage means; extracting photographing information from each of the selected image files; calculating the distribution range of the extracted photographing information; determining, based on the calculated result, a similar range of the photographing information; matching a keyword to the determined similar range of the photographing information; storing the key word and the similar range; determining whether photographing information of an image file stored in the image storage means is within the similar range or not; and assigning, in a case where the photographing information is determined to be in the similar range, the corresponding keyword to the image file. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image search technique, and more particularly to a technique for searching for a desired image including an image to which no keyword is assigned.

  As a method for specifying a search key in a conventional image search, a method is known in which a keyword is assigned to an image and an image is extracted as a result of a text search using the keyword.

  Japanese Patent Application Laid-Open No. 2004-228561 discloses a technique for making an image search process more efficient by grouping keywords according to meaning and further prioritizing when performing an image search using a plurality of keywords.

In the technique disclosed in Patent Document 2, the first keyword is generated by using at least one of the input date and time and the name of the image data, and the second keyword is generated by referring to information related to the first keyword. By using these first and second keywords, it is possible to assign keywords that are easy for the operator to understand.
JP 2000-29901 A JP-A-8-16615

  However, in the image search method described in Patent Document 1 or Patent Document 2, since an image and a keyword are associated based on human subjectivity, the same keyword is not always assigned to the same image. Therefore, there is a problem that the search result desired by the searcher is not always obtained. Furthermore, a problem has been pointed out that an image that is not associated with a keyword or the like cannot be obtained as a search result.

  The present invention has been made in view of such circumstances, and an image search program, an image search method, and an image search that can extract a desired image as a search result, including images that are not associated with keywords or the like. An object is to provide an apparatus and a recording medium.

  In order to solve the above problems, an image search program according to claim 1 according to the present invention includes an image storage means for storing an image file including an image and photographing information of the image, and an image input for reading the image file from the outside. An image search program that is executed by an image management apparatus having a search means for searching for a desired image file from an image file stored in the image storage means, and the image stored in the image storage means Image selection step for selecting a plurality of files, shooting information extraction step for extracting shooting information from each image file selected in the image selection step, and calculation step for calculating a distribution range of shooting information extracted in the shooting information extraction step And a range determination step for determining a similar range of the photographing information based on the result calculated in the calculation step. A keyword corresponding step for associating a keyword with the similar range of the photographing information determined in the range determining step, a keyword storing step for storing the keyword and the similar range of the photographing information associated with the keyword, A determination step for determining whether shooting information of the image falls within a similar range corresponding to the keyword with respect to the image file stored in the image storage means, and the shooting information is the keyword in the determination step If it is determined that it is in the similar range corresponding to the keyword stored in the storing step, the computer is caused to execute a keyword adding step for adding the keyword to the image file.

  According to a second aspect of the present invention, there is provided the image search program according to the second aspect of the present invention, wherein the keyword display step displays a list of keywords stored in the keyword storage step, and the keyword display. The method further includes a keyword selection step for selecting a keyword displayed in the step, and a search step for searching for an image file to which the keyword selected in the keyword selection step is assigned.

  The image search program according to claim 3 of the present invention is the image search program according to the invention described above, wherein the second image selection step of selecting an image file stored in the image storage means, A keyword specifying step for specifying a keyword assigned to the image file selected in the second image selection step; and a search step for searching for an image file assigned with the same keyword as the keyword specified in the keyword specifying step. .

  According to a fourth aspect of the present invention, there is provided the image search program according to the fourth aspect of the present invention, wherein each time the image file is read by the image input means, the determination step and It further includes a management information storing step of executing the keyword adding step and storing the assigned keyword and its image file in the storage means.

  The image search program according to claim 5 of the present invention is the image search program according to the above-described invention, wherein the shooting information is information stored in an Exif format image file.

  According to a sixth aspect of the present invention, there is provided a recording medium according to the present invention, an image storage unit that stores an image file including an image and photographing information of the image, an image input unit that reads an image file from the outside, and the image storage unit. A recording medium on which an image search program to be executed by an image management apparatus having search means for searching for a desired image file from stored image files is recorded, and a plurality of image files stored in the image storage means An image selection step to select, a shooting information extraction step to extract shooting information from each image file selected in the image selection step, a calculation step to calculate a distribution range of the shooting information extracted in the shooting information extraction step; A range determining step for determining a similar range of the photographing information based on the result calculated in the calculating step, and the range A keyword corresponding step for associating a keyword with the similar range of the photographing information determined in the determining step, a keyword storing step for storing the keyword and the similar range of the photographing information associated with the keyword, and the image storage means With respect to the stored image file, a determination step for determining whether shooting information of the image falls within a similar range corresponding to the keyword, and at the determination step, the shooting information is stored at the keyword storage step. When it is determined that the image is in the similar range corresponding to the keyword, a program for causing the computer to execute a keyword adding step for adding the keyword to the image file is recorded.

  The image search method according to claim 7 of the present invention includes an image storage unit that stores an image file including an image and photographing information of the image, an image input unit that reads an image file from outside, and the image storage unit. An image search method for an image management apparatus, comprising: a search unit that searches for a desired image file from image files stored in the image file, wherein the image selection step selects a plurality of image files stored in the image storage unit; A shooting information extraction step for extracting shooting information from each image file selected in the image selection step, a calculation step for calculating a distribution range of shooting information extracted in the shooting information extraction step, and a result calculated in the calculation step Based on the above, a range determining step for determining a similar range of the shooting information, and the shooting determined in the range determining step A keyword correspondence step for associating a keyword with a similar range of information, a keyword storage step for storing the keyword and a similar range of the photographing information associated with the keyword, and an image file stored in the image storage means On the other hand, in the determination step for determining whether or not the shooting information of the image falls within the similar range corresponding to the keyword, and in the determination step, the similarity in which the shooting information corresponds to the keyword stored in the keyword storage step And a keyword assigning step for assigning the keyword to the image file when it is determined that the image file is within the range.

  According to an eighth aspect of the present invention, there is provided an image retrieval apparatus according to the present invention, wherein an image storage means for storing an image and an image file including photographing information of the image, an image input means for reading an image file from the outside, and the image storage means An image search device having a search means for searching for a desired image file from the image files stored in the image storage means, the image selection means for selecting a plurality of image files stored in the image storage means, and the image selection Based on the result calculated by the calculation means, the shooting information extraction means for extracting shooting information from each image file selected by the means, the calculation means for calculating the distribution range of the shooting information extracted by the shooting information extraction means, A range determining means for determining a similar range of the photographing information, and a keyword for associating a keyword with the similar range of the photographing information determined by the range determining means Response means, keyword storage means for storing a similarity range of the keyword and the shooting information associated with the keyword, and image information stored in the image storage means for the shooting information of the image. When the determination unit determines whether or not the similar range corresponding to the keyword is included, and the determination unit determines that the shooting information is within the similar range corresponding to the keyword stored in the keyword storage unit And a keyword assigning means for assigning the keyword to the image file.

  According to the present invention, it is possible to extract a desired image as a search result including an image that is not associated with a keyword or the like.

[First Embodiment]
In the image retrieval method according to the embodiment of the present invention, image classification / management / retrieval is executed based on information (hereinafter referred to as “photographing information”) given when an image is photographed.

  FIG. 1 is a diagram for explaining the principle of image classification and retrieval. In order to simplify the description, two-dimensional coordinates are taken as an example.

  The X axis and Y axis in FIG. 1 represent shooting information (for example, shutter speed and shooting date and time). Each point shown in the figure represents a photographed image. From this figure, it can be inferred that the photographed images can be classified into groups according to the photographing information. Therefore, it is possible to search for a desired image by determining which group the image to be searched belongs to and extracting images belonging to the corresponding group.

  FIG. 2 is a diagram illustrating an example of attached information of an Exif format image file which is a kind of shooting information. These pieces of information are attached to the image file and associated with the image. Therefore, it is considered that such information is implicitly given as keywords to the image without making the user aware of it.

  Based on the above considerations, the present invention performs image search processing by treating image information as an implicit keyword and applying mathematical processing.

  FIG. 3 is a diagram showing a configuration of the image search apparatus 10 to which the image search method according to the embodiment of the present invention is applied.

  The image search apparatus 10 includes a central processing unit 11, a display unit 12, and an input / output unit 13. The central processing unit 11 executes image classification / management / retrieval processing. The display unit 12 is a CRT or TFT liquid crystal display that displays a screen related to image search. The input / output unit 13 includes an input device such as a keyboard and a mouse for receiving an operation instruction input from a user, and a communication interface that exchanges information such as image data with an external device.

  The central processing unit 11 is provided with a processing unit 15 that performs processing such as image search and a storage unit 16 that stores information related to image search processing.

  The processing unit 15 includes a photographing information extraction unit 20, a distribution / range determination unit 21, an image search unit 22, a keyword selection 23, and a keyword search unit 24. The shooting information extraction unit 20 extracts image information from the identified image file. The distribution / range determination unit 21 generates image distribution information using the shooting information as a parameter, and determines a range belonging to a specific group from the distribution information. The keyword selection unit 22 selects a specific keyword from the keyword list. The keyword search unit 23 specifies an image file having a specific keyword. The image search unit 24 searches for a specific image file.

  The storage unit 16 includes an image storage unit 27, a keyword storage unit 28, and a management data storage unit 29. The image storage unit 27 stores an image file. The keyword storage unit 28 stores the keyword and the distribution range in association with each other. The management data storage unit 29 stores keywords and shooting information in association with each other.

  Subsequently, the procedure of the image search process according to the embodiment of the present invention will be described.

  The user executes registration processing in the keyword storage unit 28 as preparation for image search. In this processing, a keyword representing a situation appropriately set by the user (hereinafter referred to as “situation keyword”) and information regarding the corresponding image file are stored in the keyword storage unit 28.

  FIG. 4 is a flowchart showing a schematic procedure of registration processing in the keyword storage unit 28.

  In step S01, the user operates and inputs a situation keyword from the input / output unit 13. The situation keyword is, for example, a word representing a situation when an image is taken, and corresponds to “indoor”, “sunset”, “night view”, and the like.

  In step S02, the user selects an image suitable for adding the situation keyword from among the image files stored in the image storage unit 27. Hereinafter, the selected image is referred to as a sample image. If the sample image is selected in a doubled manner (S03 Yes), an error message is output to the display unit 12 in step S04 and waits until the user selects the next image.

  If the sample image has not been doubled and selected (S03 No), the input situation keyword is associated with the sample image in step S05. If a sample image is further selected (S06 No), the processing from step S02 is repeated. Note that an appropriate number of sample images may be selected according to the user's intention.

  When the selection is completed (S06 Yes), the photographing information extraction unit 20 extracts Exif information in the image file of the sample image. Then, the analysis / range determination unit 21 executes the similarity analysis process (FIG. 5) shown in step S08.

  The similarity analysis process is a process of calculating the Mahalanobis distance based on the Exif information of the sample image and determining a range similar to the sample image. Specifically, when a white dot in the center of FIG. 1 represents a sample image, a range including data of the white dot (hereinafter referred to as “sample data”), that is, FIG. Then, the range drawn by the ellipse is determined by calculation. Hereinafter, the Mahalanobis distance will be described.

Considering a general case, it is assumed that there are p variables that can be considered as factors. At this time, generalized distance D ² of Mahalanobis in the formula (1).

In the equation (1), (x−μ) represents a difference between the data x ₁ , x ₂ ,... Of the variable 1, the variable ₂ ,... And the population average μ ₁ , μ ₂ ,. When expressed by each element, it is expressed by equation (2).

  In Expression (1), (x−μ) ′ is a matrix formed by replacing the rows and columns of (x−μ), that is, a transposed matrix, and is represented by Expression (3).

  In equation (1), Σ is a variance-covariance matrix and is represented by equation (4).

Here, σ _i ² (i = 1, 2,..., P), which is a diagonal element of the variance-covariance matrix, is the variance of the variable i,
σ _ij (i, j = 1, 2,..., p where i and j are different) represents the covariance between the variable i and the variable j. Σ ⁻¹ is an inverse matrix of Σ.

  The procedure of the similarity analysis process shown in FIG. 5 shows a procedure for realizing the above-described calculation.

In step T01 of FIG. 5, the average value of the sample data, that is, the population average (μ) is calculated. In step T02, the variance (σ _i ² ) of the sample data is calculated. In step T03, the covariance (σ _ij ) of the sample data is calculated. In step T04, a variance-covariance matrix (Σ) of sample data is calculated. In step T05, the inverse matrix (Σ ⁻¹ ) of the variance-covariance matrix of the sample data is calculated. In step T06, the Mahalanobis distance (D ² ) is calculated based on the average value and the inverse matrix of the variance-covariance matrix, and the process returns.

Returning to FIG. 4, in step S09, the distribution range is obtained. Here, the distribution range for one variable is the range of the X axis, which is one axis corresponding to the range depicted by the ellipse shown in FIG. And this distribution range expresses "center point" and "width" as one set. The “center point” uses the population mean μ of the variable. The “width” uses the width calculated for the variable based on the Mahalanobis distance (D ² ) described above.

  Then, the obtained distribution range and situation keyword are registered in the keyword storage unit 28. FIG. 7 is a diagram illustrating a configuration of the keyword storage unit 28. In the keyword storage unit 28, the distribution range of each piece of shooting information (each Exif information) is recorded for each situation keyword input by the user. According to the data stored in the keyword storage unit 28, for example, it can be seen that the Exif information X2 of an image having the situation keyword “sunset” is distributed in a range where the center point is 170 and the width is 170. This distribution range represents a similar range for the situation keyword.

  After the user has registered the situation keyword in the keyword storage unit 28, the registration process of the management data storage unit 29 is executed based on this registration information. This process is a process of assigning a situation keyword for each image for all images stored in the image storage unit 27.

  FIG. 8 is a flowchart showing a schematic procedure of registration processing in the management data storage unit 29.

  In step S11, a situation keyword loop is started. That is, the following processing is repeated for all situation keywords for each situation keyword. In step S12, an all-image loop is started. That is, the following processing is repeatedly executed for all images for each image.

  In step S13, the keyword storage unit 28 is searched to extract a similar range of shooting information for one situation keyword selected from the keyword storage unit 28. In step S14, shooting information stored in the image file of the image to be processed is extracted.

  In step S15, it is checked whether or not the image to be processed is within the similar range of the situation keyword. That is, it is checked whether each piece of shooting information of the image to be processed belongs to the distribution range (similar range) of the shooting information stored in the keyword storage unit 28, and the similar range of all the shooting information to be processed is included. If it belongs (S15 Yes), the image and the situation keyword are associated with each other and temporarily stored.

  If the similarity determination process is repeated for all images in step S17, the all image loop is terminated. In step S18, the temporarily stored image and situation keyword are stored in the management data storage unit 29. FIG. 9 is a diagram illustrating a configuration of the management data storage unit 29. The management data storage unit 29 includes an “image unique ID” that is information for specifying the image, an “image reference address” that indicates the position in the image storage unit 27 where the image is stored, and a thumbnail image of the image. “Index image” and at least one situation keyword associated with the image are provided.

  Then, when the above-described processing is executed for all the situation keywords stored in the keyword storage unit 28, the situation keyword loop is terminated in step S19.

  The registration process in the management data storage unit 29 described above can be additionally performed for newly read images. For example, an image read from a camera or the like via the input / output unit 13 can be additionally registered in the management data storage unit 29.

  FIG. 10 is a flowchart showing a schematic procedure of additional registration processing in the management data storage unit 29.

  In step S21, an image input loop is started. That is, the following processing is executed for each image read via the input / output unit 13. In step S22, a situation keyword loop is started. That is, the following processing is repeated for all situation keywords for each situation keyword.

  In step S23, shooting information stored in the image file of the input image to be processed is extracted. Next, in step S24, the keyword storage unit 28 is searched to extract a similar range of shooting information corresponding to the selected situation keyword.

  In step S25, it is checked whether or not the image to be processed is within the similar range of the situation keyword. That is, it is checked whether each piece of shooting information of the image to be processed belongs to the distribution range (similar range) of the shooting information stored in the keyword storage unit 28, and the similar range of all the shooting information to be processed is included. When it belongs (S25 Yes), in step S26, the situation keyword is stored in the management data storage unit 29 for the input image.

  Then, when the above-described processing is executed for all the situation keywords stored in the keyword storage unit 28, the situation keyword loop is terminated in step S27. In step S28, when the similarity determination process is repeated for all the images to be read, the image input loop is terminated.

  After the registration process of the keyword storage unit 28 and the registration process of the management data storage unit 29 are completed, the image can be searched. In the image search method according to the present embodiment, a method for searching by specifying a keyword and a method for searching for an image similar to the image by specifying an image can be used.

  FIG. 11 is a flowchart showing a schematic procedure of image search processing using keywords.

  When the user performs an image search execution operation using keywords via the input / output unit 13, a list of registered situation keywords is displayed on the display unit 12 in step S31. When the user selects the displayed situation keyword, the management data storage unit 29 is searched in step S32 to extract all images having the selected situation keyword. In step S33, thumbnail images of the extracted images are displayed as a search result list.

  FIG. 12 is a flowchart showing a schematic procedure of image search processing based on an image.

  In step S36, when the user selects an image file in the image storage unit 27 via the input / output unit 13 and instructs to search for an image similar to this image, the selection is made from the management data storage unit 29 in step S37. All situation keywords associated with the selected image are extracted.

  Next, in step S38, the management data storage unit 29 is searched to extract all images having the selected situation keyword. In step S39, thumbnail images of the extracted images are displayed as a search result list.

[Effect of the embodiment]
In the image processing method according to the present embodiment, each image is represented in a feature space based on image information (Exif information) at the time of shooting associated with the image, and a pre-selected distribution range of the sample image is used as a Mahalanobis distance. And register it in the keyword storage unit together with the situation keyword. Next, it is determined whether each Exif information of all the images that can be managed falls within this distribution range. Then, the situation keyword is registered in the management data storage unit of the image determined to enter.

  Then, the situation keywords registered in the keyword storage unit are displayed as a list, the management data storage unit is searched using the selected situation keyword as a key, and the search results are displayed as a list.

  In addition, an image file is selected from the image storage unit, a situation keyword registered in the management data storage unit of the selected image is extracted, the management data storage unit is searched using this as a key, and the search results are displayed in a list.

  Further, when the Exif information of the input image is extracted when the image data is input from the camera or the like and it is determined whether or not it is within the distribution range of each keyword in the keyword storage unit, the situation is stored in the management data storage unit. Register keywords.

  As a result of the above configuration, images taken under the shooting conditions desired by the searcher, including images that are not subject to keyword assignment or association based on human subjectivity to the image data itself, are used as search results. It can be extracted.

  In addition, since a situation keyword that accurately represents the shooting state of each image is assigned to the management data storage unit of the image input into the apparatus by the image input means without being assigned by a person, the person again assigns the keyword. There is no need.

  Therefore, it is possible to perform a search that is clearly understandable for the searcher rather than searching directly using the value of the shooting information included in the image data as a search condition.

  In the above-described embodiment, the distribution range of the preselected sample image is calculated using the Mahalanobis distance. However, the distribution range calculation method is not limited to the Mahalanobis distance, and is defined based on a statistical method. The distribution range may be determined based on each index to be used, for example, an average value, variance, maximum value, minimum value, and the like.

  Each function described in the above embodiment may be configured using hardware, or may be realized by reading a program describing each function into a computer using software. Each function may be configured by appropriately selecting either software or hardware.

  Furthermore, each function can be realized by causing a computer to read a program stored in a recording medium (not shown). Here, as long as the recording medium in the present embodiment can record a program and can be read by a computer, the recording format may be any form.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. Further, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, you may combine suitably the component covering different embodiment.

The figure explaining the principle of classification and search of an image. The figure which shows the attached information of the image file of an Exif format. The figure which shows the structure of the image search apparatus with which the image search method of embodiment of this invention is applied. The flowchart which shows the general | schematic procedure of the registration process to a keyword memory | storage part. The flowchart which shows the outline procedure of a similarity analysis process. The figure which shows the distribution range about one variable. The figure which shows the structure of a keyword memory | storage part. The flowchart which shows the general | schematic procedure of the registration process to a management data storage part. The figure which shows the structure of a management data storage part. The flowchart which shows the general | schematic procedure of the additional registration process to a management data storage part. The flowchart which shows the general | schematic procedure of the image search process by a keyword. The flowchart which shows the general | schematic procedure of the image search process based on an image.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Image retrieval apparatus, 11 ... Central processing unit, 12 ... Display part, 13 ... Input-output part, 15 ... Processing part, 16 ... Memory | storage part, 17 ... Processing part, 20 ... Imaging | photography information extraction part, 21 ... Analysis and range Determining unit, 22 ... keyword selection unit, 23 ... keyword search unit, 24 ... image search unit, 27 ... image storage unit, 28 ... keyword storage unit, 29 ... management data storage unit.

Claims (8)

Image storage means for storing an image and an image file including shooting information of the image;
An image input means for reading an image file from the outside;
An image search program to be executed by an image management apparatus having search means for searching for a desired image file from image files stored in the image storage means,
An image selection step of selecting a plurality of image files stored in the image storage means;
A shooting information extraction step of extracting shooting information from each image file selected in the image selection step;
A calculation step of calculating a distribution range of the shooting information extracted in the shooting information extraction step;
Based on the result calculated in the calculating step, a range determining step for determining a similar range of the photographing information;
A keyword correspondence step for associating a keyword with the similar range of the shooting information determined in the range determination step;
A keyword storing step for storing a similarity range of the photographing information associated with the keyword and the keyword;
A determination step of determining whether or not shooting information of the image falls within a similar range corresponding to the keyword with respect to the image file stored in the image storage unit;
If it is determined in the determining step that the shooting information is within a similar range corresponding to the keyword stored in the keyword storing step, a keyword adding step for adding the keyword to the image file A program to make it run. A keyword display step for displaying a list of keywords stored in the keyword storage step;
A keyword selection step for selecting a keyword displayed in the keyword display step;
The program according to claim 1, further comprising a search step of searching for an image file to which the keyword selected in the keyword selection step is assigned. A second image selection step of selecting an image file stored in the image storage means;
A keyword specifying step for specifying a keyword assigned to the image file selected in the second image selection step;
The program according to claim 1, further comprising: a search step of searching for an image file to which the same keyword as the keyword specified in the keyword specifying step is assigned.   Management information storage for executing the determination step and the keyword assigning step on the image file each time the image input means reads the image file, and storing the assigned keyword and the image file in the storage means The program according to claim 1, further comprising a step.   The program according to claim 1, wherein the shooting information is information stored in an Exif format image file. Image storage means for storing an image and an image file including shooting information of the image;
An image input means for reading an image file from the outside;
A recording medium on which an image search program to be executed by an image management apparatus having search means for searching for a desired image file from image files stored in the image storage means is recorded;
An image selection step of selecting a plurality of image files stored in the image storage means;
A shooting information extraction step of extracting shooting information from each image file selected in the image selection step;
A calculation step of calculating a distribution range of the shooting information extracted in the shooting information extraction step;
Based on the result calculated in the calculating step, a range determining step for determining a similar range of the photographing information;
A keyword correspondence step for associating a keyword with the similar range of the shooting information determined in the range determination step;
A keyword storing step for storing a similarity range of the photographing information associated with the keyword and the keyword;
A determination step of determining whether or not shooting information of the image falls within a similar range corresponding to the keyword with respect to the image file stored in the image storage unit;
If it is determined in the determining step that the shooting information is within a similar range corresponding to the keyword stored in the keyword storing step, a keyword adding step for adding the keyword to the image file A recording medium on which a program to be executed is recorded. Image storage means for storing an image and an image file including shooting information of the image;
An image input means for reading an image file from the outside;
An image search method for an image management apparatus comprising search means for searching for a desired image file from image files stored in the image storage means,
An image selection step of selecting a plurality of image files stored in the image storage means;
A shooting information extraction step of extracting shooting information from each image file selected in the image selection step;
A calculation step of calculating a distribution range of the shooting information extracted in the shooting information extraction step;
Based on the result calculated in the calculating step, a range determining step for determining a similar range of the photographing information;
A keyword correspondence step for associating a keyword with the similar range of the shooting information determined in the range determination step;
A keyword storing step for storing a similarity range of the photographing information associated with the keyword and the keyword;
A determination step of determining whether or not shooting information of the image falls within a similar range corresponding to the keyword with respect to the image file stored in the image storage unit;
And a keyword assigning step of assigning the keyword to the image file when the determining step determines that the shooting information is in a similar range corresponding to the keyword stored in the keyword storing step. An image search method characterized by that. Image storage means for storing an image and an image file including shooting information of the image;
An image input means for reading an image file from the outside;
An image search device having search means for searching for a desired image file from image files stored in the image storage means,
Image selection means for selecting a plurality of image files stored in the image storage means;
Shooting information extraction means for extracting shooting information from each image file selected by the image selection means;
Calculating means for calculating a distribution range of the photographing information extracted by the photographing information extracting means;
Based on the result calculated by the calculating means, range determining means for determining a similar range of the photographing information;
Keyword correspondence means for associating a keyword with the similar range of the shooting information determined by the range determination means;
Keyword storage means for storing a similarity range of the photographing information associated with the keyword and the keyword;
A determination unit that determines whether or not shooting information of the image falls within a similar range corresponding to the keyword with respect to the image file stored in the image storage unit;
When the determination unit determines that the shooting information is within a similar range corresponding to the keyword stored in the keyword storage unit, the determination unit includes a keyword adding unit that adds the keyword to the image file. An image search device characterized by the above.

Images