JP2003303188A - System and method for presenting similar image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To present an image similar to an image selected by a user on a user terminal having an access through a network. <P>SOLUTION: This similar image presenting system comprises a means for receiving a retrieval request of the similar image specifying an auction ID from a user computer 4, a means for obtaining characteristic data corresponded to the auction ID specified by the retrieval request from a characteristic database 7 having recorded characteristic data of each article image. The system comprises a means for retrieving the characteristic database 7 and extracting the auction ID corresponded to the characteristic data similar to the extracted characteristic data, a means for obtaining image data corresponding to the extracted auction ID from an image database 5, and a means for transmitting the obtained image data to the user computer 4. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a similar image presentation system and method for retrieving an image similar to an image designated by a user terminal accessed through a network and browsing the image with the user terminal. .

[0002]

2. Description of the Related Art In an auction or shopping on the Internet, a product image is generally displayed to a user together with a description of the product content. in this case,
One of the needs of the user is a function of presenting a product that has a similar appearance when a certain product is selected. For example, in the case of products such as clothes and ornaments, color,
The appearance of shapes, patterns, etc. can be a great motivation for the product. Therefore, if other products similar to the appearance of the product being browsed by the user can be displayed on the auction site or the shopping site, it is effective in increasing the user's willingness to purchase.

[0003]

However, the above-described effect of presenting a similar image has not been recognized so far, and another product similar in appearance to a certain product is displayed on a site on the Internet. The service was not provided. The present invention has been made in view of the above points, and an object thereof is to present an image similar to an image designated by a user to a user terminal that has accessed via a network.

[0004]

In order to achieve the above object, the invention described in claim 1 searches for an image similar to an image designated by a user terminal that has accessed through a network. In the similar image presentation system used for browsing at the user terminal, an image database for storing image data of various images and an image ID for each image data stored in the image database
And a feature database in which feature data representing features of the image are recorded in association with each other, and the image I
Means for receiving a similar image search request specifying D, characteristic data acquisition means for acquiring characteristic data associated with the image ID specified in the received search request from the characteristic database, and the characteristic database And a similar image search means for extracting an image ID associated with the feature data similar to the extracted feature data,
Image data acquisition means for acquiring image data corresponding to the extracted image ID from the image database, and image data transmission means for transmitting the acquired image data to the user terminal are provided.

According to the first aspect of the invention, the characteristic data representing the characteristic of each image is stored in the characteristic database.
Therefore, it is possible to determine the degree of similarity between images by comparing these characteristic data. Therefore, according to the present invention, an image similar to the image designated by the user can be presented.

According to the invention described in claim 2, in the system according to claim 1, the characteristic data is divided and recorded according to the category of the article represented by the corresponding image. To do.

The invention described in claim 3 is the system according to claim 2, wherein the similar image search means specifies a corresponding category from the image ID, and characteristic data belonging to the specified category. Is a target of the search.

According to the third aspect of the invention, since the characteristic data belonging to a common category may be searched in the characteristic database, the search target is narrowed down, and the search can be performed quickly.

Further, the invention according to claim 4 searches for an image similar to the image designated by the user terminal that has accessed through the network, and presents a similar image to be browsed by the user terminal. It relates to the method.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention applied to an auction system on a network will be described below. In an Internet auction, a user who bids a bidder for a given period of time on the Internet for a given item and offers the highest bid price has the right to purchase the item for sale at the bid price from the seller. To get it. The system of this embodiment is
It has a function of searching an image similar to the image specified by the user for the product image provided by the seller of the auction product and presenting it to the user.

FIG. 1 is an overall configuration diagram of the system of this embodiment. As shown in FIG. 1, the system according to the present embodiment includes an auction server 1 and a feature server 2. The feature server 2 is connected to the auction server 1. In addition, the auction server 1 is connected to the user computer 4 via the network 3.
Are connected.

As will be described later, the auction server 1
When a user issues a search request for a similar image while browsing an image of a product, the user transmits an auction ID unique to the product to the feature server 2. When the characteristic server 2 receives the auction ID from the auction server 1,
A product image similar to the image corresponding to the auction ID is searched, and the auction ID corresponding to the searched similar image is transmitted to the auction server 1.

As shown in FIG. 1, the auction server 1
Has an image database 5 and an auction database 6. Further, the feature server 2 is provided with a feature database 7 and a category database 8.

The image database 5 stores image data representing each commodity put up for auction. In addition, the auction database 6 stores auction data including a bid price, the number of bids, and the remaining time of the auction for each product. The feature database 7 also stores feature data representing the features of each product image. Further, the category database 8 stores category data indicating the correspondence relationship between the auction ID and the category of the corresponding product.

FIG. 2 shows an example of auction data recorded in the auction database 6. As shown in the figure, in the auction database 6, auction data is recorded for each category. The auction data is composed of data items such as an auction ID, a URL of image data of a product image, a bid price in the auction, the number of bids, and the remaining time of the auction. The “current price” in FIG. 2 indicates, for example, the price of the bidder who bid the second highest price, plus the amount of money for the bid unit.

FIG. 3 shows an example of the characteristic data recorded in the characteristic database 7. As shown in the figure, the characteristic data is also recorded by being classified according to the category of the product, like the auction data. In the present embodiment, the color, edge (contour), and texture (pattern or pattern) are extracted from the image data of all products provided by the seller of the product to be sold at the auction by the feature extraction function of the feature server 2 described later. Three types of features that are patterns) are extracted, and the data is stored in the feature database 7 as feature data.

Next, the function of the characteristic server 2 will be described. The characteristic server 2 has the following two functions. (1) From the image data, color, edge (outline),
And a function (feature extraction function) of extracting three types of features of texture (pattern or pattern) to obtain feature data. (2) When there is a request from the auction server 1 to search for a similar image in which the auction ID is specified, the auction ID of the image similar to the image of the specified auction ID is extracted and the data indicating the high degree of similarity is obtained. Together with the function of notifying the auction server 1 (similarity judgment function). Below is a brief description of these features (see
See, for example, Japanese Patent Laid-Open No. 2000-285141).

<Feature Extraction Function> There are three types of features extracted by the feature extraction function: the color of the product, the edge representing the outline of the product, and the texture (pattern or pattern) on the product surface.

When extracting the color feature of the product image,
First, the color of each pixel included in the image is red, green,
And the coordinates of the three-dimensional color vector space based on the RGB color system specified by the gradation values of blue and blue are specified. An RGB histogram formed by performing such correspondence for all pixels is
Used as color feature data for product images.

In the case of extracting the edge feature of the product image, the entire image data is converted into an array of numbers which is the primary difference by the Sobel operator or the like. Generally, in the array area corresponding to the edge of the image, the first-order difference value shows a large value. Therefore, it is considered that an edge exists near the array where the primary difference value is large.
That is, it can be said that the array of the first-order difference values digitizes the features of the edges of the image. Therefore, the primary difference value array by the Sobel operator is used as the edge feature data of the product image.

In the case of extracting the texture of the product image, the power spectrum obtained by Fourier-transforming the data in which the brightness of the image is digitized and arrayed is the direction of the pattern or pattern of the brightness of the image and the image element. Can be said to describe the distribution of the size of. Therefore, the power spectrum in the array obtained by converting the array of pixels, which is the two-dimensional coordinate of the image, into the complex plane is used as the texture feature data of the product image.

The color feature data, edge feature data, and texture feature data described above are all represented as vector data. In the present embodiment, when registering a product to be newly put up, the feature data of the image of the product is extracted by the feature extraction function of the feature server 2 and recorded in the feature database 7. .

<Similarity judgment function> Auction server 1
If there is a search request for a similar image that specifies the auction ID, the feature server 2 first refers to the category correspondence table described above and identifies the category of the feature database 7 from the specified auction ID. Next, the characteristic server 2 reads the characteristic data of the specified auction ID, and the Euclidean distance between this characteristic data and each characteristic data belonging to the specified category is calculated as color characteristic data, edge characteristic data, and These three types of Euclidean distances are added together to obtain each of the texture feature data. Here, the Euclidean distance is defined as the distance between the vector end points of the two feature data which are vector data. Therefore, it is determined that the smaller the total of the three types of Euclidean distances is, the higher the overall similarity is in the characteristics of color, edge, and texture.

In the present embodiment, it is assumed that the images are similar when the calculated Euclidean distance is smaller than a predetermined reference value. That is, the feature server 2 extracts an auction ID in which the total of the three types of Euclidean distances is smaller than the reference value, and together with the auction ID, the total value of the three types of Euclidean distances indicates the degree of similarity. The data is transmitted to the auction server 1 as data (hereinafter referred to as similarity data). The similarity is not limited to the above Euclidean distance, and it is also possible to define a unique distance according to the type of image feature data and use it as the similarity data.

Next, the flow of processing from when the user accesses the auction server 1 to make a search request for similar images until the similar images are displayed on the user computer 4 will be described with reference to FIG. .

First, when the user computer 4 accesses the auction server 1 (S100), the auction server 1 displays a category list of auction items on the user computer 4 (S101).
When the user computer 4 selects a category (S1
02), the auction server 1 displays a list of products belonging to the selected category on the user computer 4 (S103). When the user selects a product from the product list (S104), the auction server 1
The product image illustrated in FIG. 5 is transmitted from the user computer 4 to the user computer 4 (S105). As shown in FIG. 5, on the product screen, a product image, a “similar item search” button 21,
A "bid" button 22 is arranged. By clicking the "bid" button 22 on this product screen, a bid can be placed on the product.

When the user wants to browse a product that is similar in appearance to the displayed product, he / she can search for a similar item in the product image.
Click button 21. In response, the user computer 4 sends to the auction server 1 the auction ID of the product being displayed (“K12” in the example of FIG. 5).
3 ”) is designated, and a similar image search request is transmitted (S).
106).

In response to the above search request, the auction server 1 sends a search request for a similar image for the specified auction ID to the feature server 2 (S10).
7).

When the feature server 2 receives the search request from the auction server 1 (S108), the feature server 2 extracts the auction ID of the product image similar to the product image of the specified auction ID by the similarity determination function (S109). . Then, the extracted auction ID is transmitted to the auction server 1 together with the similarity data (S110).

If there is no auction ID having similarity data smaller than the reference value in the feature data,
The feature server 2 notifies the auction server 1 that there is no similar image (S111).

When the auction server 1 receives the auction ID and the similarity data corresponding to the auction ID from the feature server 2 (S112), the auction database 6 is referred to, and from the auction data of the category specified by the auction ID, URL of image data corresponding to each received auction ID
Is acquired (S113). Next, the image data of the acquired URL is read by referring to the image database 5 (S
114). Then, together with the original product image, a retrieval result display screen that displays the read image in the ascending order of the similarity data value (that is, in the descending order of similarity) is generated and transmitted to the user computer 4 (S115). This screen is displayed on the user computer 4 (S116).

When the feature server 2 notifies that there is no similar image, a screen including a message that there is no similar product is transmitted to the user computer 4 (S117).

FIG. 6 is a diagram showing an example of the search result display screen displayed on the user computer 4 in S116. In the example shown in the figure, the original product image requested by the user to search for a similar image is arranged at the top of the search result display screen. Below this image, the category name, the total number of similar product images searched, and
A list of similar product images is displayed. By clicking each image, it is possible to move to the display screen of each product.
As described above, the display order of similar images is the order with the highest degree of similarity by default, but by clicking "current price", "bid number", and "remaining time" included in the screen, respectively You can sort by price, number of bids, and remaining time.

As described above, according to the system of the present embodiment, the user browses the data of other products that are similar in appearance to the put-up product, and refers to the bid price when the user participates in the auction. Alternatively, it is possible to find a product closer to one's own taste and bid.

Further, since the feature data is divided into categories and recorded, and the feature server 2 uses the feature data of the corresponding category as a search target for similar images, the search target can be narrowed down to improve the search speed. it can. However, if the feature server 2 has a sufficiently high-speed processing capability, the feature data does not necessarily have to be divided into categories.

In the above embodiment, the case where the present invention is applied to the auction system has been described.
The present invention is not limited to this, and can be provided to a system that presents various images to the user, such as an electronic mall system.

[0037]

According to the present invention, an image similar to the image selected by the user can be presented to the user terminal that has accessed via the network.

[Brief description of drawings]

FIG. 1 is a diagram showing an overall configuration of a system according to an embodiment.

FIG. 2 is a diagram showing an example of auction data in the present embodiment.

FIG. 3 is a diagram showing an example of feature data in the present embodiment.

FIG. 4 is a diagram showing a flow of processing in the present embodiment.

FIG. 5 is a diagram showing an example of a product display screen displayed on a screen of a user computer.

FIG. 6 is a diagram showing an example of a search result display screen displayed on the screen of the user computer.

[Explanation of symbols]

1 Image server 2 feature server 3 network 4 User computer 5 Image database 6 auction database 7 feature database 8 category database

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) G06F 17/60 ZEC G06F 17/60 ZEC (72) Inventor Masajiro Iwasaki 1-3 Nakamagome, Ota-ku, Tokyo -6 F-term in Ricoh Co., Ltd. (reference) 5B075 KK33 KK37 ND06 QM05

Claims (4)

[Claims] 1. A similar image presentation system which searches for an image similar to an image designated by a user terminal that has accessed through a network and provides the image for browsing at the user terminal, the image data of various images. An image database in which the image ID is stored, a feature database in which, for each image data stored in the image database, an image ID and feature data representing the feature of the image are recorded in association with each other; A means for receiving a search request for a similar image that specifies the feature data, a feature data acquisition means for acquiring feature data associated with the image ID specified by the received search request from the feature database, and the feature database. A similar image for searching and extracting an image ID associated with feature data similar to the extracted feature data A search means; an image data acquisition means for acquiring image data corresponding to the extracted image ID from the image database; and an image data transmission means for transmitting the acquired image data to the user terminal. And the system. 2. The similar image presentation system according to claim 1, wherein the characteristic data is recorded by being divided according to the category of the article represented by the corresponding image. 3. The similar image retrieving means includes the image ID.
3. The similar image presentation system according to claim 2, wherein the corresponding category is specified from the above, and the feature data belonging to the specified category is set as a search target. 4. A similar image presentation method of searching for an image similar to an image designated by a user terminal that has accessed via a network and using the image for browsing at the user terminal, the image being displayed from the user terminal. A procedure for receiving a search request for a similar image specifying an ID, and a feature in which the image ID and feature data representing the feature of the image are recorded in association with each other for each image stored in the image database that stores various images. A procedure for acquiring characteristic data associated with the image ID specified by the received search request from the database, and searching the characteristic database for association with characteristic data similar to the extracted characteristic data. A step of extracting an image ID, a step of acquiring image data corresponding to the extracted image ID from the image database, Method characterized by comprising the step of transmitting the obtained image data to the user terminal.