JP2019008597A - Image search method and server - Google Patents

Abstract

To provide an image search method and a server that classify keywords attached to each image into group tags and individual tags and group an appropriate number of images so as to make it easy to locate a desired image.SOLUTION: An image search method includes the steps of: detecting at least one associated keyword for each of a plurality of images; aggregating the number of times of detection of each keyword; grouping images in which keywords whose number of times of detection is equal to or more than a first predetermined value and equal to or less than a second predetermined value are associated; transmitting data in an image browsing page including the plurality of images and a group tag in which the number of times of detection is a keyword equal to or more than the first predetermined value and equal to or less than the second predetermined value; and searching the plurality of grouped images when the group tag in the image browsing page is selected.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an image search method and a server.

  A service is provided in which an image taken by a customer with a digital camera, a smartphone, or the like is uploaded to a print order receiving system, and a photo print order can be ordered (see, for example, Patent Document 1).

  In addition, a service for selling photographs taken by professional photographers at events such as concerts and sports competitions is known (see, for example, Patent Document 2). Image data taken by a professional photographer is uploaded to a photo sales system via a network. The customer can browse the uploaded image and select and purchase a favorite photo. The purchased image is printed at a printing shop, and the printed matter is delivered to the customer.

  Customers need to find out images of their favorite singers and players, images of themselves and their families, etc. from the large number of images posted in the photo sales system. It was.

JP 2014-142711 A JP 2001-325082 A

  The present invention has been made in view of the above-described conventional situation, and the keywords attached to each image are classified into group tags and individual tags to be handled separately, and an appropriate number of images are grouped to obtain a desired It is an object of the present invention to provide an image search method and a server capable of easily searching for images.

  An image search method according to the present invention includes a step of detecting at least one associated keyword for each of a plurality of images, a step of counting the number of times of detection of each keyword, and the number of times of detection being equal to or greater than a first predetermined value. (2) grouping images associated with keywords having a predetermined value equal to or smaller than a predetermined value, the plurality of images, and a group tag whose number of detections is a keyword equal to or larger than the first predetermined value and equal to or smaller than the second predetermined value. A step of transmitting image browsing page data; and a step of searching the plurality of grouped images when the group tag in the image browsing page is selected.

  In the image search method according to an aspect of the present invention, when a keyword is input via the image browsing page, the image associated with the group tag that matches the input keyword and the number of detections that match the input keyword are At least one of images associated with an individual tag that is a keyword less than the first predetermined value is searched.

  In the image search method according to an aspect of the present invention, keywords having the number of detections greater than the second predetermined value among the keywords associated with each image are classified as keywords that are not used for deletion or search.

  The server according to the present invention transmits image data of a plurality of images to an analysis terminal and receives an analysis result including at least one keyword associated with each image from the analysis terminal; and each of the plurality of images And a totaling unit that detects the associated keywords and counts the number of detections of each keyword, and the plurality of images and a group tag that is a keyword whose number of detections is greater than or equal to a first predetermined value and less than or equal to a second predetermined value. A generating unit that generates an image browsing page, and when the group tag in the image browsing page is selected, the generating unit searches for a plurality of images associated with the keyword of the group tag, An image browsing page including a plurality of extracted images is generated.

  In the server according to one aspect of the present invention, when a keyword is input via the image browsing page, the generation unit matches the image associated with the group tag that matches the input keyword and the input keyword. A search is made for at least one of images associated with individual tags, which are keywords whose detection count is less than the first predetermined value.

  The server according to an aspect of the present invention classifies, as keywords that are not used for deletion or search, keywords that are detected more than the second predetermined value among keywords associated with each image.

  According to the present invention, it is possible to easily search for a desired image by classifying keywords assigned to each image into a group tag and an individual tag to divide the handling and grouping an appropriate number of images.

1 is a schematic diagram of an image sales system according to an embodiment of the present invention. It is a schematic block diagram of the server which concerns on the same embodiment. It is a functional block diagram implement | achieved by running a search processing program. It is a flowchart explaining the image search method by the embodiment. (A)-(e) is a figure which shows the example of the keyword provided to an image. (A)-(c) is a figure which shows the example of an image browsing page. It is an external view of the image printing apparatus by another embodiment.

  FIG. 1 is a schematic diagram of an image sales system according to an embodiment of the present invention. The image sales system includes a server 1 that distributes a sales image browsing page to a user terminal 4 of a customer, a registration terminal 2 that is used by a right holder to register a sales image, an analysis that analyzes sales images, and assigns keywords. The terminal 3 is included.

  When a customer selects a favorite image from a browsing page using a user terminal 4 such as a smartphone or a PC and performs a settlement process, the customer downloads image data of a sales image or orders a printed matter. Can do. When the server 1 accepts the order for the printed matter, it transmits the image data to the factory 5. The printed matter produced at the factory 5 is delivered to the customer. The printed matter is, for example, a photograph, a poster, a photo book, a mug, or the like.

  As shown in FIG. 2, the server 1 is a computer having a CPU 10 (central processing unit), a storage unit 11, a communication unit 12, and the like. The server 1 may further include a display unit 13 such as a liquid crystal display and an operation unit 14 such as a keyboard and a mouse.

  The communication unit 12 performs communication processing with the registration terminal 2, the analysis terminal 3, the user terminal 4, and the like, and transmits and receives data. The communication unit 12 receives image data to be sold from the registration terminal 2 and stores it in the storage unit 11. The communication unit 12 transmits the image data of the sales image to the analysis terminal 3, receives keyword information related to each image from the analysis terminal 3, and stores the keyword information in the storage unit 11. The image analysis process will be described later.

  The communication unit 12 generates a sales image browsing page and transmits it to the user terminal 4. Further, the communication unit 12 receives an image selection result and payment information from the user terminal 4.

  The storage unit 11 is a combination of a flash memory, a ROM (Read-Only Memory), a RAM (Random Access Memory), and the like. Image data of sales images, keyword information given to each image, classification of keyword classification processing described later As a result, a search processing program and the like are stored.

  When the CPU 10 executes the search processing program, as shown in FIG. 3, a totaling unit 16, a classification unit 17, and a browsing page generation unit 18 are realized. The totaling unit 16 detects the keywords assigned to each image by the analysis terminal 3 and totals the number of detections for each keyword.

  The classification unit 17 determines (classifies) each keyword as a group tag, an individual tag, or an unnecessary tag based on the number of times of detection. The group tag is a tag for grouping a plurality of images. The individual tag is a tag indicating the characteristics of each image. A keyword assigned in common to a large number of images cannot be used for searching / extracting the images (the images cannot be narrowed down and has no meaning to use), and thus is classified as “unnecessary”.

  The browsing page generation unit 18 generates a sales image browsing page using a group tag or an individual tag.

  The image search method according to the present embodiment will be described along with a flowchart shown in FIG. 4 together with a specific example.

  The right holder of the image registers the sales image in the server 1 using the registration terminal 2 (step S1). Sales image data is stored in the storage unit 11 of the server 1.

  The server 1 transmits the image data of the sales image to the analysis terminal 3, and receives the analysis result including the keyword assigned to each image from the analysis terminal 3 (step S2). The analysis terminal 3 performs image recognition by a known method, and associates (assigns) a keyword related to the image. For example, the analysis terminal 3 detects an object included in the image, classifies it from many categories such as flowers, animals, and vehicles, and assigns them as keywords. Further, the analysis terminal 3 recognizes and extracts text included in the image by OCR (optical character recognition), and assigns it as a keyword.

  The analysis terminal 3 may detect a person's face shown in the image, detect an emotional state, an attribute such as wearing a hat, and the like, and assign it as a keyword. Further, the analysis terminal 3 may detect a famous product or company logo included in the image, detect a famous natural structure or artificial structure, and give the detection result as a keyword.

  The server 1 may provide the analysis terminal 3 with reference data in which a person image and a person name are associated with the sales image. The analysis terminal 3 may compare the face image detected from the sales image with the reference data, specify a person in the image, and give the person name as a keyword.

  Examples of images transmitted to the analysis terminal 3 and assigned keywords are shown in FIGS.

  The totaling unit 16 detects the keywords assigned to each image and totals the number of detections (step S3). For example, the number of detection times of keywords assigned to the images shown in FIGS. 5A to 5E is as shown in Table 1 below.

  The classification unit 17 classifies each keyword as a group tag, an individual tag, or unnecessary based on the counting result. The classification unit 17 selects one unselected keyword from the keywords detected in step S3 (step S4).

  The number of times of detection of the selected keyword is compared with the first predetermined value N1 and the second predetermined value N2. The second predetermined value N2 is larger than the first predetermined value N1. The administrator of the server 1 can arbitrarily set the predetermined values N1 and N2. When the number of times of detection is not less than the first predetermined value N1 and not more than the second predetermined value N2 (step S5_Yes, step S6_No), this keyword is classified into a group tag (step S7). In the present embodiment, keywords classified into group tags are further classified into individual tags.

  When the number of times of detection is less than the first predetermined value N1 (step S5_No), this keyword is classified into individual tags (step S8). If the number of times of detection is greater than the second predetermined value N2, this keyword is classified as “unnecessary” (step S9). The classification unit 17 may delete a keyword classified as “unnecessary” from keywords assigned to each image, or may classify the keyword as a keyword that is not used for the search. Whether a keyword classified as “unnecessary” is deleted or classified as a keyword that is not used for the search can be determined according to the load on the server 1 or the like.

  The above classification process is performed for all keywords (steps S4 to S10). The classification result is stored in the storage unit 11 (step S11).

  For example, in the examples shown in FIGS. 5A to 5E and Table 1, if the first predetermined value N1 is 2 and the second predetermined value N2 is 4, each keyword is classified as shown in Table 2 below. The

  The keywords “summer” and “rice ball” that are detected twice or more and four times or less are suitable for grouping a plurality of images, and therefore are set as group tags. For example, the image shown in FIG. 5A and the image shown in FIG. 5C are grouped by a group tag “summer”. Similarly, the image shown in FIG. 5B and the image shown in FIG. 5E are grouped by the group tag “rice ball”.

  The keywords “drink”, “ramune”, “shaved ice”, “peach blossom”, and “Hinamatsuri”, which are detected once, are characteristic information that distinguishes images from others, and are therefore individual tags.

  The keyword “food” with more than 4 detections is common to 5 images, and when grouped with this keyword, all images will be grouped. Not suitable. Further, it is not characteristic information that distinguishes an image from others. Therefore, this keyword is classified as “unnecessary”.

  The browsing page generation unit 18 generates a browsing page for selling an image using the keyword classification result (step S12). The browsing page displays sales images individually and also displays group tags. When the customer operates the user terminal 4 to select a group tag, the browsing page generation unit 18 searches and extracts images grouped by this group tag and displays them.

  The browsing page has a keyword input field. When the customer operates the user terminal 4 to input a keyword, the browsing page generation unit 18 searches for, extracts, and displays an image having the input keyword as an individual tag.

  For example, the browsing page generation unit 18 generates a browsing page 80 as shown in FIG. 6A and transmits it to the user terminal 4. On the browsing page 80, images 81 to be sold are individually displayed, and a purchase button 82 and a keyword input field 84 for each image 81 are displayed. In addition, group tags 83 “summer” and “rice ball” are displayed on the browsing page 80.

  For example, when the group tag 83 “Summer” is selected, the browsing page generation unit 18 searches for an image assigned with “Summer” as a keyword, and generates a browsing page as shown in FIG. .

  Further, for example, when the keyword “ramune” is input in the keyword input field 84, the browsing page generation unit 18 searches for an image assigned with the keyword “ramune”, as shown in FIG. 6C. Generate a browsing page.

  The customer searches for a favorite image using narrowing down by group tag, keyword search, etc., and presses the purchase button 82 to select a purchase image (step S13). When the selection of the image is finished, a settlement process is performed (step S14). After the settlement process, the server 1 transmits image data to the user terminal 4 and transfers the contents of the print order to the factory 5.

  As described above, according to the present embodiment, the number of detections (number of appearances) of the keywords assigned by the analysis terminal 3 is tabulated, and keywords whose number of detections is not less than the first predetermined value N1 and not more than the second predetermined value N2 are grouped. Since a tag is used, an appropriate number of images can be grouped to make it easier to find an image. Therefore, it is possible to reduce the trouble for the customer to search for a desired image from a large number of images.

  In the above embodiment, the display / non-display of the group tag 83 in the browsing page 80 may be arbitrarily set. For example, a photo seller (server administrator) selects a group tag to be displayed from among a plurality of group tags. The browsing page generation unit 18 generates a browsing page on which the selected group tag is displayed. The photo seller may be able to specify the display order of the group tags.

  The customer may be allowed to switch display / non-display of the group tag. In this case, the browsing page 80 is provided with a switching button for switching display / non-display of the group tag 83.

  For example, when the number of group tags is large (a predetermined number or more), the browsing page generation unit 18 first generates a browsing page in which the group tag is not displayed. When the customer presses the switching button to turn on the display, the group tag is displayed. When the number of group tags is a small number (below a predetermined number), the browsing page generation unit 18 may first generate a browsing page displaying the group tags.

  In the said embodiment, although the example of classification | category of the keyword provided to the food image was shown, a sales image is not limited to this, It can apply to various images.

  For example, in the case of a marathon tournament photo, the analysis terminal 3 extracts numbers (bib numbers) and team names displayed on the bib numbers by OCR and assigns them as keywords. Further, the analysis terminal 3 gives the keyword “marathon” to a large number of images.

  The classification unit 17 of the server 1 classifies the keyword “marathon” having a large number of detections as “unnecessary”. The classification unit 17 classifies team names whose detection count is equal to or greater than the first predetermined value N1 and equal to or smaller than the second predetermined value N2 as group tags. On the browsing page, the team name is displayed as a group tag. When a customer who is a participant in a marathon event selects a group tag of the team name to which he belongs, an image showing himself or his teammate is retrieved and displayed.

  When the sales image is a photograph of a concert of an idol group, the server 1 provides the analysis terminal 3 with reference data in which each member's face image and name are associated with each other. The analysis terminal 3 compares the face image detected from the image with the reference data, specifies a member in the image, and assigns the name as a keyword. In addition, the analysis terminal 3 gives keywords such as “idol” and “singer” to a large number of images.

  The classifying unit 17 of the server 1 classifies the keywords “idle” and “singer” that are frequently detected as “unnecessary”. The classification unit 17 classifies the member names whose detection count is equal to or greater than the first predetermined value N1 and equal to or smaller than the second predetermined value N2 as group tags. The member name is displayed as a group tag on the browsing page. When a customer who is a fan of an idol group selects a group tag having a favorite member name, an image showing this member is retrieved and displayed.

  In the above embodiment, when a keyword is input on the browsing page, an example in which an image having the input keyword as an individual tag is searched for, extracted, and displayed has been described. However, even if an image having the input keyword as a group tag is searched for Alternatively, an image having the input keyword as a group tag or an individual tag may be searched.

  In the above-described embodiment, an example in which a print page manufactured by the factory 5 is delivered when a browsing page is transmitted to the user terminal 4 and a print order is received has been described, but an image as a kiosk terminal as illustrated in FIG. The sales image may be registered in the printing apparatus 7 and the printed matter may be sold.

  Touch panels 70 </ b> A and 70 </ b> B are provided on the top plate 71 </ b> A portion and the upper part of the casing 71 of the image printing apparatus 7, respectively. A housing 71 accommodates a high-resolution color printer such as a sublimation type or an ink jet type. A transparent opening / closing door 73 is provided on the front side of the housing 71 so as to be freely opened and closed. On the back surface side of the opening / closing door 73, there is provided an accumulating portion (not shown) in which the printed matter printed out by the printer is accumulated. It has become.

  The image printing apparatus 7 also includes a billing unit including a coin insertion slot 72A, a bill insertion slot 72B, and a change return slot 72C provided on the front side of the casing 71, and is inserted from the coin insertion slot 72A and the bill insertion slot 72B. The received money is identified and counted, and the change is discharged from the change return port 72C.

  The image printing apparatus 7 is connected to the server 1 so as to be communicable, and receives image data to which keywords are assigned by the analysis terminal 3 from the server 1. The image printing apparatus 7 has a computer and can execute the above-described search processing program and perform the same processing as the server 1. The image printing apparatus 7 generates a browsing page as described above and displays it on the touch panel 10A (or 10B).

  When the customer operates the touch panel 10A to select a purchase image, a billing process is performed. After the accounting process, the image printing apparatus 7 prints the selected image on photographic paper and outputs it. The customer can receive the printed matter on the spot.

  The search processing program described in the above-described embodiment may be stored in a recording medium such as a CD-ROM and read by a computer and executed. The recording medium is not limited to a removable medium such as a magnetic disk or an optical disk, but may be a fixed recording medium such as a hard disk device or a memory. The search processing program may be distributed via a communication line (including wireless communication) such as the Internet. Further, the program may be distributed in a state where the program is encrypted, modulated or compressed, and stored in a recording medium via a wired line such as the Internet or a wireless line.

  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. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

1 server 2 registration terminal 3 analysis terminal 4 user terminal 5 factory 6 printed matter 7 image printing device

Claims (6)

Detecting at least one associated keyword for each of the plurality of images;
A step of counting the number of detections for each keyword;
Grouping images associated with keywords whose number of detections is greater than or equal to a first predetermined value and less than or equal to a second predetermined value;
Transmitting data of the plurality of images and an image browsing page including a group tag whose number of detections is a keyword that is not less than the first predetermined value and not more than the second predetermined value;
When the group tag in the image browsing page is selected, searching the plurality of grouped images;
An image search method comprising:   When a keyword is input through the image browsing page, an image associated with the group tag that matches the input keyword, and an individual that has the number of detections that matches the input keyword is less than the first predetermined value The image search method according to claim 1, wherein at least one of images associated with a tag is searched.   The image search method according to claim 1 or 2, wherein among keywords associated with each image, a keyword having the number of detection times greater than the second predetermined value is classified as a keyword that is not used for deletion or search. . A communication unit that transmits image data of a plurality of images to an analysis terminal, and receives an analysis result including at least one keyword associated with each image from the analysis terminal;
For each of the plurality of images, an associated keyword is detected, and a counting unit that counts the number of times each keyword is detected;
A generating unit configured to generate an image browsing page including the plurality of images and a group tag whose number of detections is a first predetermined value or more and a second predetermined value or less;
With
When the group tag in the image browsing page is selected, the generation unit searches for a plurality of images associated with the keyword of the group tag and generates an image browsing page including the extracted plurality of images. A server characterized by that.   When a keyword is input via the image browsing page, the generation unit includes an image associated with the group tag that matches the input keyword, and the number of detections that matches the input keyword is less than the first predetermined value. The server according to claim 4, wherein at least one of images associated with an individual tag that is a keyword is searched. 6. The server according to claim 4, wherein among keywords associated with each image, a keyword having the number of detection times greater than the second predetermined value is classified as a keyword that is not used for deletion or search.

Images