WO2001024050A1

WO2001024050A1 - Image recognition system and database system for image recognition

Info

Publication number: WO2001024050A1
Application number: PCT/JP2000/003637
Authority: WO
Inventors: Toshiaki Nagai
Original assignee: Cadix Inc.
Priority date: 1999-09-27
Filing date: 2000-06-05
Publication date: 2001-04-05
Also published as: JP2001101191A; AU4954400A

Abstract

Images of works of art are recognized to identify their attribute information such as the name of the artists who produced the works. A digital camera (10) takes pictures of a painting (12) and generates the digital images of the painting. Feature extraction means (14) extracts the features of the digital images of the painting. Retrieve means (16) compares the determined features with the features stored in a database (18) to find the image information corresponding to the features of the picture. The retrieval means (16) reads the attribute information about the image, such as the title and the painter, from the database (18), and sends it to output means (20), which in turn supplies it to display means (22) to present the title, the painter, and so on. Users can thus find the titles of paintings and the names of painters.

Description

Description Image identification device and database system used for image identification

The present invention relates to an image identification device. In particular, the present invention relates to an image identification device capable of acquiring an image of a painting or the like and knowing the creator of the painting. The present invention also relates to a recording medium having a data structure used by the image identification device, and a database system used by the image identification device.

Background art

In recent years, the so-called digital lure power eve business of preserving artworks such as paintings with digital information has been widely performed. By storing artwork in digital information in this way, unlike the real thing, it can be stored semi-permanently. By creating a database of such artworks, for example, it will be possible to instantly display the artwork from the title of the work, which is expected to contribute to education and research on artworks. Disclosure of the invention

According to the database of works of art from such digital archiving business, works of art can be searched instantly based on the title of the work, the name of the creator, etc. It was difficult to know the title of the work, that is, to reverse the database.

In addition, even in a plant picture book etc., even though there is an index for searching for a picture of a plant from the name of the plant, it was difficult to know the name of the plant from the image (picture) of the plant.

For example, if one wants to know the names of plant varieties that exist in the field, and if there is a system that can identify the varieties of the plants from the plant images, it can easily be expected that this will greatly contribute to education and research. It is not hard to imagine that if the name of the work and the creator are known based on the images of the paintings, etc., the art will greatly contribute to education and research.

However, conventionally, there is no known system or database that can know the attributes (variety name, creator name) of the object (plant, beauty product) represented by the image information based on the image information.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and has as its object to provide a system capable of knowing attribute information such as the creator of an object represented by image information based on the image information of a work of art or the like. That is.

According to a first aspect of the present invention, in order to solve the above-described problems, a feature amount calculating unit that calculates a feature amount of given input image information, and image information having a feature amount that matches the calculated feature amount is stored in a database. And an attribute information output unit that outputs attribute information of the object represented by the image information found by the search from the database, and an attribute information output unit that outputs the read attribute information. It is an image identification device.

According to such a configuration, since the search is performed using the feature amount, the attribute information of the object represented by the image information can be efficiently obtained.

According to a second aspect of the present invention, there is provided a database storing image information, attribute information of an object represented by the image information, and a feature amount of the image information, and a feature amount of a given input image information is calculated. An amount calculating unit, and a searching unit that searches the database for image information having a feature amount matching the calculated feature amount, and reads out attribute information of an object represented by the image information found by the search from the database. And an attribute information output means for outputting the read attribute information.

According to such a configuration, since the database is included, there is no need to access an external database.

According to a third aspect of the present invention, there is provided an image identification apparatus comprising: digital camera means for generating image information of an object; wherein the digital camera means generates the input image information.

Because it has digital camera means to generate an image of the object,

7 Image information can be quickly generated based on an object, and attribute information of the object can be quickly obtained.

A fourth aspect of the present invention is the image identification device, wherein the object is a work of art, and the attribute information includes at least an author name and a work name of the work of art.

3

According to such a configuration, the name of the creator of the artwork can be quickly known.

A fifth aspect of the present invention is the image identification device, wherein the object is a painting, and the attribute information includes at least a creator name and a work name of the painting.

According to such a configuration, the name of the painting or the like can be quickly known.

A sixth aspect of the present invention is the image identification device, wherein the object is clothing, and the attribute information includes at least a brand name of the clothing.

According to such a configuration, the brand name and the like of the clothes can be quickly known. A seventh aspect of the present invention is the image identification device, wherein the object is a tie, and the image information is image information representing a pattern of the nectar.

According to such a configuration, the brand name of the tie can be quickly known based on the tie pattern.

An eighth aspect of the present invention is the image identification device, wherein the object is a plant, and the attribute information includes at least a name and a classification of the plant. According to such a configuration, the name of the plant and the like can be quickly known.

A ninth aspect of the present invention is the image identification device, wherein the object is an animal, and the attribute information includes at least a name and a classification of the animal. According to such a configuration, the name and the like of the animal can be quickly known.

A tenth aspect of the present invention is the image identification device, wherein the object is a fish, and the attribute information includes at least a name and a classification of the fish. According to such a configuration, the name and the like of the fish can be quickly known.

In the eleventh aspect of the present invention, the feature amount includes at least a freshness feature amount emphasizing an eye color of the fish, and the database includes the freshness feature An image identification apparatus comprising a plurality of entries having different amounts. According to such a configuration, the freshness of fish can be identified.

A twelfth aspect of the present invention is the image identification device, wherein the object is a food, and the attribute information includes at least a name of the food.

According to such a configuration, the freshness of the food can be identified.

According to a thirteenth aspect of the present invention, a feature amount calculating means for calculating a feature amount of given input image information, and image information having a feature amount matching the calculated feature amount are searched from a database. An image identification device comprising: a search unit that reads attribute information in a state represented by image information found by the above from the database; and an attribute information output unit that outputs the read attribute information.

According to such a configuration, since the search is performed using the feature amount, the attribute information of the state represented by the image information can be efficiently obtained.

According to a fourteenth aspect of the present invention, a database for storing image information, attribute information of a state represented by the image information, a feature amount of the image information, and a feature amount of given input image information is calculated. A feature amount calculation unit, and image information having a feature amount matching the calculated feature amount is searched from the database, and attribute information of a state represented by the image information found by the search is retrieved from the database. An image identification device comprising: a retrieval unit that reads out; and an attribute information output unit that outputs the attribute information that is read out.

According to such a configuration, a database is included, so it is not necessary to access an external database.

A fifteenth aspect of the present invention is an image discriminating apparatus comprising: a digital camera means for generating image information representing a predetermined state; wherein the digital camera means generates the input image information.

Since the digital camera is provided to generate an image representing the state, image information can be quickly generated based on the state in front of the eyes, and attribute information of the state can be obtained quickly. .

A sixteenth aspect of the present invention is characterized in that the state is a state of a ground surface, a road surface, or a floor surface representing footprints of shoes, and the attribute information includes at least a name of a manufacturer of the shoes. Image identification device. According to such a configuration, it is possible to quickly know the name of the shoe manufacturer with respect to the footprint of the shoe.

A seventeenth aspect of the present invention is the image identification device, wherein the state is a medical examination result, and the attribute information includes at least a diagnosis result.

According to such a configuration, the diagnosis result can be quickly known based on the medical examination result.

An eighteenth aspect of the present invention is the image identification device, wherein the image information is image information representing a lentogen photograph.

According to such a configuration, the diagnosis result can be quickly known based on the lentogen photograph.

A ninth aspect of the present invention is a storage unit for storing image information representing an object, attribute information of the object represented by the image information, and a feature amount of the image information, and According to such a database system, a search means for searching for image information, and an attribute of an object represented by the image information are obtained by a query including externally input image information. We can know information. According to a twenty-second aspect of the present invention, there is provided storage means for storing: image information representing a state; attribute information of the state represented by the image information; and a feature amount of the image information. A database system comprising: a search unit for searching for image information; and

According to such a database system, it is possible to know the attribute information of the state represented by the image information by an inquiry including the externally input image information. A twenty-first aspect of the present invention is a recording medium storing image information representing an object, attribute information of the object represented by the image information, and a feature amount of the image information. A computer-readable recording medium characterized in that the image information can be searched using a feature amount as a key.

According to the storage medium having such a data structure, the image information can be searched based on the feature amount of the image information, and further, the attribute information of the object represented by the image information can be obtained. A twenty-second aspect of the present invention is a recording medium characterized by storing image information representing a state, attribute information of the state represented by the image information, and a feature amount of the image information. A computer-readable recording medium characterized in that the image information can be searched using a feature amount as a key.

According to the storage medium having such a data structure, the image information can be searched based on the feature amount of the image information, and further, the attribute information of the state represented by the image information can be obtained.

According to a twenty-third aspect of the present invention, there is provided a recording medium reading means for reading the input image information from a predetermined recording medium, wherein the characteristic amount calculating means calculates the characteristic amount of the input image information read from the recording medium. An image identification apparatus characterized in that:

According to such a configuration, an image identification device capable of identifying an image can be obtained based on the recording medium on which the image is stored.

A twenty-fourth aspect of the present invention is the image identification device, wherein the recording medium is a recording medium on which image information can be written by digital camera means.

According to the second configuration, it is possible to first capture an image with a digital camera, store the captured image on a recording medium, and then perform image identification based on the recording medium on which the image is stored. .

A twenty-fifth aspect of the present invention includes: an image receiving unit that receives the human-powered image information from a predetermined communication line.The feature amount calculating unit calculates a feature amount of the input image information received by the image receiving unit. An image identification device characterized in that it is calculated.

According to the second configuration, it is possible to perform image identification on input image information transmitted via a communication line.

A twenty-sixth aspect of the present invention is an image identification device, comprising: attribute information transmitting means for transmitting the attribute information output by the attribute information output means via the communication line.

According to such a configuration, the attribute information as a result of the image identification can be transmitted via the communication line, and the image identification device can be used from a remote place.

In a twenty-seventh aspect of the present invention, the image receiving means receives an electronic mail from the communication line. And an e-mail receiving means for extracting the input image information contained in the e-mail.

According to such a configuration, it is possible to perform image identification on the input image information included in the e-mail.

In a twenty-eighth aspect of the present invention, the image transmitting means includes an electronic mail transmitting means for transmitting, via the communication line, an e-mail including the attribute information output by the attribute information output means. This is a featured image identification device.

According to such a configuration, the attribute information as a result of the image identification can be transmitted by e-mail, and the image identification device can be used from a remote place.

A twentieth aspect of the present invention is the image identification device, wherein the communication line is the Internet.

Since the Internet was used as the communication line, the image identification device can be used from a place where it can access the Internet.

A thirty-fifth aspect of the present invention is the mobile terminal receiving means for receiving the input image information from a digital camera means via the mobile communication line, wherein the communication line is a mobile communication line. An image identification device comprising: According to such a configuration, image identification can be performed based on input image information received via the mobile communication line.

A thirty first aspect of the present invention is the mobile communication line, wherein the communication line is a mobile communication line for performing communication by a mobile communication terminal, and the image transmitting means transmits the attribute information output by the attribute information output means to the mobile communication terminal. A mobile terminal transmitting means for transmitting the mobile communication terminal to the other mobile communication terminal via a line.

According to such a configuration, attribute information as a result of image identification can be received by the mobile communication terminal, and attribute information generated by the image identification device can be known from a remote location.

A thirty-second aspect of the present invention is the image identification device, wherein the mobile communication line is a mobile phone communication line.

Since input image information and attribute information are transmitted and received using the mobile phone communication line, the image identification device can be used from a place where a mobile phone can be used. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a block diagram of a configuration of an image identification device according to a preferred embodiment of the present invention.

FIG. 2 is a configuration block diagram of a database system according to a preferred embodiment of the present invention.

FIG. 3 is an explanatory diagram showing the data structure of the database.

FIG. 4 is a block diagram of a configuration of an image identification device according to another embodiment. BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.

Embodiment 1

FIG. 1 shows a configuration block diagram of an image identification device 8 according to a preferred first embodiment of the present invention.

As shown in this figure, the image identification device 8 is composed of a feature value calculating means 14, a searching means 16, and an output means 20.

First, the digital camera 10 photographs the painting 12 and generates a digital image of the painting. The generated digital image is sent to the feature amount calculation means 14. The feature amount calculating means 14 calculates the feature amount of the digital image of the painting.

Next, the search means 16 compares the calculated feature quantity with the feature quantity stored in the database 18 and finds image information corresponding to the matching feature quantity. This image information is the image information of the painting.

Further, the search means 16 reads out the author name, the name of the picture, etc., which are the attribute information of the found image information, from the database 18 and sends them to the output means 20. The output means 20 causes the display means 22 to display the supplied author name, painting name, and the like. In this way, according to the first embodiment, by photographing a painting with the digital camera 10, the creator of the painting, the name of the painting, and the like are indicated on the display means 22. Therefore, the user can easily know the name of the painting and the name of the creator.

A characteristic of the first embodiment is that image information is stored in a database 18. That is, the feature amount is used when searching from. Various conventionally known feature amounts can be used as the image feature amount.

The characteristic amount calculating means 14 calculates such various characteristic amounts. Since the feature amount itself is conventionally known, the description of the calculation method is omitted.

Database 18 is a database related to painting. This database 18 stores image information of a painting, attribute information of the painting, and feature values of the image information. The attribute information is the author name of the painting, the name of the painting, and the like. In addition, the database 18 is provided with an index table related to the characteristic amount in order to make it easy to search the image information of the painting using the characteristic amount as a key. As a result, a data structure capable of quickly searching for image information using the feature amount as a key is realized in the database 18.

In the first embodiment, the search unit 16 reads the attribute information from the database 18 with respect to the image information searched using the feature amount as a key. Then, the search means 16 supplies the read attribute information, that is, the author name and the like to the display means 22. The display means 22 displays the author name and the like.

Database 18 can take various forms. For example, it may be configured on a hard disk of a computer. Also, for example, it is considered that this kind of painting database is often set up in museums and the like. In such a case, the image identification device 8 is often located far from the database 18. In this case, to adopt a mode of connection to the database 1-8 via a communication line image identification apparatus 8 is preferably _c of database system suitable for connecting the image identification apparatus 8 via such communication channel The configuration is described in further detail in Embodiment 3 below.

As attribute information, it is also preferable to use, in addition to the author's name and the name of the painting, the year of the painting's production, the owner, the size, and a commentary by a specialist. By displaying the commentary and explanation by the expert as attribute information on the display means 22, the user can see the explanation of the picture and contribute to the purpose of learning and education.

Also, in the above explanation, the digital image of the painting is generated by the digital camera 10. However, it is also preferable to obtain a digital image of a painting by scanning a picture of the painting with a scanner. In addition, a configuration may be adopted in which a still image of a painting is captured from a video signal. In any case, any method can be adopted as long as a digital image of the painting can be obtained.

In the above description, the database 18 is configured separately from the image identification device 8, but may be included in the image identification device 8. When the types of paintings to be identified are small and the database 18 is small, it is preferable to configure the database 18 on a hard disk or the like inside the image identification device.

Various display devices can be used for the display device 22. Various display devices such as a CRT and a liquid crystal display device that have been conventionally used can be used. Furthermore, in the above description, the display device 22 is configured separately from the image identification device 8, but may be included inside the image identification device 8. In this case, if a small liquid crystal display device is provided in the image identification device 8, there is a possibility that a miniaturized image identification device that is convenient to carry can be realized. It is preferable that the feature amount calculation means 14 use hardware dedicated to calculation, but it is also preferable to realize the feature amount software. Also, it is desirable that the search means 16 be constituted by software for accessing the database 18. Further, the output means depends on the type of the external display means 22, but it is desirable that the output means be constituted by, for example, a video power and software for driving the video power.

Embodiment 2 (application fields)

In the above description, examples of paintings have been described. However, the objects are not limited to paintings, but may be other artworks, animals, plants, clothes, etc., as long as they can be photographed with a digital camera or the like.

twenty one

Sculptures can be considered as works of art in addition to paintings. In this case, the database 18 is a database including sculpture image information, feature amounts of sculpture image information, and sculpture attribute information (creator of the sculpture, name of the sculpture).

twenty two

When the object is a plant, it can be used as a plant picture book that allows the name and type of the plant to be known based on the image of the plant. In this case, the database 18 is a database containing plant image information, feature amounts of plant image information, and plant attribute information. It is a base. The attribute information of a plant includes the name of the plant, scientific name, family, ecology, breeding area, annual or two-year grass, and so on.

twenty three

When the object is an animal, it is almost the same as a plant. Based on the image information of the animal, the name and ecology of the animal can be known.

If the animal is a edible fish, it is preferable to use a feature that emphasizes the color of the fish's eyes as a feature. This is based on the widespread use of fish eye color as an indicator of fish freshness. In such a case, the database 18 has a plurality of entries with different freshness for one kind of fish. For example, for image information of a fish with high freshness, the entry database includes image information of a fish with high freshness, the feature amount of image information of a fish with high freshness, and attribute information of a fish including information of high freshness. Stored in 18.

For image information of a fish with low freshness, the entry includes image information of a fish with low freshness, the feature amount of image information of a fish with low freshness, and attribute information of a fish including information of low freshness. Are stored in the database 18.

In this way, for a single kind of object, a plurality of entries with different freshness are included in the database 18. By constructing the database 18 in this way, it is possible to know not only the name of the fish but also the freshness of the fish based on the image information of the fish.

twenty four

In the above 2.3, the image identification device capable of knowing the freshness of fish has been described. However, it is also preferable to configure an image identification device capable of knowing the freshness of general food using the same principle.

In this case, as in the case of fish, the database 18 has a plurality of entries with different freshness for one type of food. For example, with respect to image information of a certain food with high freshness, the entry includes image information of a certain food with high freshness, feature amounts of image information of a certain food with high freshness, and attribute information of a certain food including information of high freshness. Is stored in database 1 8 ₌ Similarly, the database 18 also stores entries for image information of a food item with low freshness.

By using such a database 18, it is possible to configure an image identification device capable of knowing the freshness of food.

twenty five

If the object is clothing, it is the same as a work of art, etc. Based on the image information of clothing, the brand name and the type of weaving of the clothing can be known. In this case, the database 18 is a database including clothing image information, feature amounts of clothing image information, and clothing attribute information. The attribute information of the garment includes the brand name of the garment, fiber type, weaving method, washing method, price, and the like.

In particular, if the garment is a tie or a scarf, it can often be identified from the design rather than its shape. Therefore, in the case of a tie, it is preferable to use a part of the tie as image information. Then, based on the image information of the symbol, the feature amount of the image and attribute information such as a tie having the symbol are stored in the database 18.

2.6

In the example described above, the attribute information of an object is obtained by using image information representing a single independent object. As long as image information can be generated, instead of the object itself, it may be a "state" such as a "trace" of the object or the state of the object. It is also good to get attribute information.

For example, footprints of an object such as a footprint can be generated by using the digital camera 10 or the like. It is also preferable to configure the image identification device 8 that can display the attribute information of the shoe that made the footprint based on the footprint. Here, the footprints are the footprints of the shoes on the ground surface and the footprints of the shoes on the street and on the floor.

As the attribute information of the shoe, it is preferable to use the name of the shoe manufacturer, the material of the shoe, and the like.

The “state of the object” is, for example, an inspection result of an X-ray photograph image or the like. Images matching the image information of lentogen photos based on their features By searching from the database i8, the attribute information of the radiograph can be displayed.

As the attribute information, a diagnosis result for the radiograph is preferable. By displaying the diagnostic results as attribute information, it is possible to configure an automatic diagnostic device as it were.

Here, radiographs have been described, but any examination result, such as an endoscopic photograph, can be used as long as it can be expressed by an image.

Embodiment 3 (Database system)

In the above description, the image identification device 8 uses the database 18 having a specific data structure. As described above, the database 18 includes, in each entry, image information of an object, feature amounts of the image information of the object, and attribute information of the object. It is. As described above, the database 18 includes, in each entry, state image information, feature amounts of the state image information, and attribute information of the state. Is also a feature. By the way, in general, a database of works of art and the like is often managed in a museum or the like. Therefore, as described above, it is considered that the image identification device 8 and the database 18 are often connected by a communication line.

In this case, rather than the database 8 as a mere storage device, preferably constructed as a database system 3 0 having also search function _e This is due to the reduction of Torafui poke amount of communication times line. A configuration block diagram of such a database system 30 is shown in FIG.

When the database system 30 is used instead of the database 18, the search means 16 in FIG. 1 issues a query (query 1) to the database system 30, and the database management means 32 calls this query. In response to (query), the database 18 is searched and the result is returned to the search means 16. According to such a configuration, since only the search result flows through the communication line, the traffic amount can be reduced. Preferably, the database management means 32 is realized by computer software:

If the database system 30 is installed in an art museum, etc., Instead of using the image identification device 8, it is also possible for the user to manually know the name of the painting or the like.

That is, the user first obtains image information by photographing a work of art using a digital camera 10 or the like. Next, the user calculates the feature amount of the image information, and requests the database system 30 via the communication line to find out the image information matching the feature amount from the database 18. Based on this request, the database management means 32 of the database system 30 searches for the corresponding image information, and transmits the name or work name, which is the attribute information of the image information, to the user. Users can learn the attributes of the artwork by receiving the name of the artwork and the author's name via the communication line.

The database system 30 having the above configuration can be used not only for a database of works of art but also for the medical diagnosis described above. In this case, the database system 30 will be installed in a hospital, for example. As a result, so-called remote diagnosis can be performed on a subject in a remote place.

Embodiment 4 (Recording medium)

In the embodiment described above, it is assumed that the database 18 having a special data structure is used. FIG. 3 shows the data structure of this database 18. As described above, this database 18 includes image information of an object, feature amounts of the image information, and attribute information of the object: FIG. However, as described above, the database 18 may be configured to include image information of a state, feature amounts of the image information, and attribute information of the state:

A database is constructed by storing such a special data structure in a recording medium. Therefore, it is possible to search for image information using the feature values. In order to speed up the search, it is also preferable to construct an index table for the feature values (see Fig. 3).

The recording medium may be any recording medium that can be read by a computer. For example, a floppy disk may be used, and it is also preferable to use a Husband CDROM. By configuring the database 18 in a computer-readable recording medium in this way, an image identification device can be easily configured using a computer. In addition, by accessing the database 18 in the recording medium by the user of the computer, it is also possible to obtain attribute information of the object or state represented by the image information based on the image information.

Embodiment 5 (Acquisition form of input image information)

In the embodiment described above, the input image information is directly supplied from an external digital camera 10 or the like as shown in FIG.

However, it is also preferable to supply image information via a communication line without directly connecting the digital camera 10 and the image identification device 8. It is also preferable to adopt a form in which image information is temporarily stored in a recording medium, and this recording medium is supplied to an image identification device.

Hereinafter, in Embodiment 5, various modes for supplying input image information to the image identification device will be described.

In FIG. 4, the image identification device 38 according to the fifth embodiment includes a feature amount calculation unit 14, a search unit 16 and an output unit 20, as in FIG. The display means 22 is connected to the output means 20, and the database 18 is connected to the search means. These configurations perform operations similar to those of the first embodiment.

5.1

What is characteristic in the fifth embodiment is that the recording medium reading means 40 is connected to the feature quantity calculating means 14. The recording medium reading means 40 reads the input image information from a recording medium such as a compact flash smart medium and supplies the input image information to the feature amount calculating means 14. The feature amount calculating means 14 calculates the feature amount based on the supplied input image information.

According to such a configuration, the input image information can be supplied to the image identification device 38 without connecting the digital camera 10 directly to the image identification device 38. That is, the user first captures an image using the digital camera 10 or the like to obtain input image information. This input image information is stored in a recording medium such as the above-mentioned compact flash: Digital cameras 10 and the like generally use such a compact flash. It has a function of storing an image in a recording medium such as a flash or a smart media. The user causes the recording medium reading means 40 of the image identification device 38 to read the recording medium storing the image in this manner. The recording medium reading means 40 supplies the read image to the feature amount calculating means 14 as input image information. The following operation is the same as in the first embodiment, and the user can see the attribute information and the like as the result of image identification on the screen of the display means 22.

By such an operation, the user can take an image at a place away from the image identification device 38.

Various types of recording media can be used as long as they are portable. In addition to the compact flash described above, a memory stick, a floppy disk, a CD-R, and the like are also preferable. Also, a recording medium such as a removable hard disk may be used.

However, in the fifth embodiment, since a recording medium is used for storing images, it is most preferable that the image acquisition means such as the digital camera 10 is a recording medium on which image information can be written.

Further, it is preferable that the recording medium reading means 40 is constituted by a drive device for reading and writing a compact flash smart medium, a memory stick, and the like, and a driver program for driving the drive device.

5.2

Furthermore, a feature of the fifth embodiment is that the e-mail communication unit 42 is connected to the feature amount calculation unit 14. The e-mail communication means 42 receives an e-mail via a communication line such as the Internet. Then, the input image information included in the e-mail is extracted and supplied to the feature amount calculating means 14. According to such a configuration, the characteristic amount calculating means 14 calculates the characteristic amount based on the supplied input image information. Even if the digital camera 10 is not directly connected to the image identification device 38, the input image Information can be provided to the image identification device 38.

That is, the user first captures an image using the digital camera 10 or the like, and stores the captured input image information in a notebook computer or the like at the location where the capture was performed. next, The user uses the communication function of a notebook computer or the like to send an e-mail with the obtained input image information attached to the image identification device.

The e-mail communication means 42 receives an e-mail transmitted via a communication line such as the Internet. Then, the input image information attached to the e-mail is extracted and supplied to the feature amount calculating means 14. The following operation is the same as that of the first embodiment, and the image identification device 38 obtains attribute information and the like as the result of image identification.

Further, the e-mail communication means 42 according to the fifth embodiment has a function of returning the obtained attribute information and the like using an e-mail. That is, the electronic mail communication means 42 returns an electronic mail including the attribute information and the like obtained by the search means 16 to the destination to which the input image information has been sent. The user can know the attribute information by looking at the returned e-mail.

As a result, the user it is possible to use the functions of the image identification apparatus 3 8 from a distance away from the image identification apparatus 3 8, _c image identification apparatus 3 8 highly convenient is realizable affirmative ϋ

Note that the image receiving means and the attribute information transmitting means of the present invention correspond to the e-mail communication means 42 of the fifth embodiment. Furthermore, the electronic mail transmitting means and the electronic mail receiving means of the present invention also correspond to the electronic mail communication means 42 of the fifth embodiment. It is preferable to comprise a modem or a terminal adapter or the like and electronic mail software.

5.3

In the example described above, an example has been described in which input image information and attribute information are transmitted and received by e-mail using the Internet. However, if the input image information and the attribute information can be transmitted and received, e-mail is not necessarily used.

In particular, in recent years, digital cameras 10 have become more sophisticated and often have a function of communicating with a remote computer using a mobile phone or the like. When such a digital camera 10 is used, it is also preferable to transmit and receive the input image information and the attribute information via a general public line such as a mobile communication line without using the Internet. In order to realize this, in the fifth embodiment, the mobile communication means 44 is connected to the feature quantity calculation means 14. The mobile communication means 44 is means for transmitting and receiving input image information and attribute information using a mobile communication line such as a mobile phone. Then, the mobile communication means 44 extracts the input image information received via the mobile communication line, and supplies it to the feature value calculation means 14. The feature value calculating means 14 calculates the feature value based on the supplied input image information.

According to such a configuration, the input image information can be supplied to the image identification device 38 without connecting the digital camera 10 directly to the image identification device 38. That is, the user first captures an image using a digital camera 10 or the like, and transfers the captured input image information to the image identification device 38 using communication via a mobile phone. Send.

The mobile communication means 44 of the image identification device 38 receives the input image information transmitted via the mobile communication line. Then, the input image information is supplied to the feature amount calculating means 14. The following operation is the same as that of the first embodiment, and the image identification device 38 obtains attribute information and the like as a result of image identification.

Further, the mobile communication means 44 according to the fifth embodiment has a function of returning the obtained attribute information and the like to the mobile phone via the mobile communication line. In other words, the mobile communication means 44 returns the attribute information and the like obtained by the search means 16 to the other party to whom the input image information was sent. By looking at the above, you can know the attribute information. In recent years, mobile phones have become more sophisticated and have functions to display not only telephone functions but also various types of information.

As a result, the user can use the functions of the image identification device 38 even from a place far from the image identification device 38, and a highly convenient image identification device 38 can be realized. The image receiving means and the attribute information transmitting means correspond to the mobile communication means 44 of the fifth embodiment. Further, it goes without saying that the mobile terminal receiving means and the mobile terminal transmitting means of the present invention also correspond to the mobile terminal communication means 44 of the fifth embodiment. Preferably, the mobile communication means 44 includes a mobile phone, an adapter for performing data communication using the mobile phone, and the like.

As described above, according to the present invention, it is possible to search for image information using the feature amount of an image, and to know attribute information of an object or a state represented by the image information. As a result, it is possible to realize an image identification device that can know the name of the work or the author of the work such as a painting based on the image information of the work such as a painting.

Also, if the image identification device is provided with a database, it is possible to know the attribute information based on the image information without accessing an external database.

Furthermore, if digital camera means is used as a means for obtaining image information, image information can be quickly generated based on an object or state in front of the eyes, and attribute information of the object or state can be quickly known.

As the object, in addition to works of art, clothes, ties, animals, plants, fish, foods, and the like can be applied. According to the present invention, attribute information of the object can be known based on these image information. it can.

In particular, in the present invention, since the freshness of fish and food can be included as attribute information, it is possible to know such freshness.

As the state, shoe prints, medical test results, and the like can be adopted, and a diagnosis result based on the name of the shoe manufacturer that produced the shoe prints and the test results can be obtained. Further, since a radiograph is used as image information, a diagnosis result can be obtained based on the radiograph according to the present invention.

In addition, since the database is provided with the database management means, even when accessing the database system via the communication line, it is possible to reduce the traffic volume of the communication line.

As described above, the database system of the present invention is provided with the database management means for retrieving the image information by the feature amount. Obtainable.

Further, since the present invention is a computer-readable recording medium storing a database having these special data structures, it is possible to search for image information from a computer using the feature amount of image information, Can also get information _C

In the present invention, the input image information is read from a predetermined recording medium by a recording medium reading means. Therefore, the attribute information of the image can be known even when the place where the image is photographed and the image identification device are far apart.

Further, according to the present invention, since means for transmitting and receiving input image information and attribute information is provided using the Internet or a mobile phone communication line, an image identification device can be used from a remote place.

Claims

The scope of the claims

1. a feature value calculating means for calculating a feature value of the given input image information;

Searching means for searching the database for image information having a feature amount matching the calculated feature amount, and reading out from the database attribute information of an object represented by the image information found by the search;

Attribute information output means for outputting the read attribute information,

An image identification device comprising:

2. A database that stores image information, attribute information of an object represented by the image information, and a feature amount of the image information.

Feature value calculating means for calculating a feature value of the given input image information;

A search unit that searches the database for image information having a feature amount that matches the calculated feature amount, and reads, from the database, attribute information of an object represented by the image information found by the search;

An image identification device comprising:

3. The image identification device according to claim 1 or 2,

Digital camera means for generating image information of an object,

An image identification device, characterized in that the digital camera means generates the input image information.

4. In the image identification device according to claim 1 or 2,

The image identification device, wherein the object is a work of art, and the attribute information includes at least an author name and a work name of the work of art.

5. The image identification device according to claim 4,

The image identification device, wherein the object is a painting, and the attribute information includes at least an author name and a work name of the painting.

6. In the image identification device according to claim 1 or 2,

The image identification device, wherein the object is clothing, and the attribute information includes at least a brand name of the clothing.

7. The image identification device according to claim 6, wherein

The image identification device, wherein the object is a tie, and the image information is image information representing a pattern of the tie.

8. The image identification device according to claim 1 or 2,

The image identification device, wherein the object is a plant, and the attribute information includes at least a name and a classification of the plant.

9. The image identification device according to claim 1 or 2,

The image identification device, wherein the object is an animal, and the attribute information includes at least a name and a classification of the animal.

10. The image identification device according to claim 9,

The image identification device, wherein the object is a fish, and the attribute information includes at least a name and a classification of the fish.

11. The image identification device according to claim 10, wherein:

The feature quantity includes at least a freshness feature quantity focused on the eye color of the fish,

The image identification device, wherein the database includes a plurality of entries having different freshness feature amounts for a certain kind of fish.

1 2. In the image identification device according to claim 1 or 2,

The image identification device, wherein the object is a food, and the attribute information includes at least a name of the food.

13. A feature amount calculating means for calculating a feature amount of the given input image information;

A search unit that searches the database for image information having a feature amount that matches the calculated feature amount, and reads out the attribute information of the state represented by the image information found by the search from the database; and the read-out attribute information. Attribute information output means for outputting

An image identification device comprising:

14. A database that stores image information, attribute information of a state represented by the image information, and a characteristic amount of the image information.

Feature value calculating means for calculating a feature value of the given input image information; A search unit that searches the database for image information having a feature amount that matches the calculated feature amount, and reads out attribute information of a state represented by the image information found by the search from the database;

An image identification device comprising:

15. The image identification device according to claim 13 or 14, wherein:

Digital camera means for generating image information representing a predetermined state,

An image discriminating apparatus, characterized in that the digital camera means generates the input image information.

16. The image identification device according to claim 13 or 14, wherein:

The image identification is characterized in that the condition is a condition of a ground surface, a road surface or a floor representing footprints of shoes, and the attribute information includes at least a name of a manufacturer of the shoes.

17. The image identification device according to claim 13 or 14, wherein:

The state is a medical examination result, and the attribute information includes at least a diagnosis result.

18. The image identification device according to claim 17,

The image identification device, wherein the image information is image information representing an X-ray photograph.

19. A storage unit for storing image information representing an object, attribute information of the object represented by the image information, and a feature amount of the image information.

Search means for searching for the image information using the feature amount as a key,

A database system comprising:

20. storage means for storing: image information representing a state; attribute information of the state represented by the image information; and a feature amount of the image information.

A database system comprising:

21. A recording medium storing image information representing an object, attribute information of the object represented by the image information, and a feature amount of the image information, wherein the feature amount is A computer-readable recording medium capable of retrieving the image information as a key.

22. A recording medium storing image information representing a state, attribute information of a state represented by the image information, and a feature amount of the image information, wherein the feature amount is set as a key, and A computer-readable recording medium capable of retrieving image information.

23. In the image identification device according to claims 1, 2, 13, or 14,

Image recording means for reading out the input image information from a predetermined recording medium, wherein the feature amount calculating means calculates a feature amount of the input image information read out from the recording medium. apparatus.

24. In the image identification device according to claim 23,

An image identification apparatus, wherein the recording medium is a recording medium on which image information can be written by digital camera means.

25. In the image identification device according to claims 1, 2, 13, or 14,

Image receiving means for receiving the input image information from a predetermined communication line,

An image identification device, comprising: a feature amount calculating unit that calculates a feature amount of the input image information received by the image receiving unit.

26. In the image identification device according to claim 25,

Attribute information transmitting means for transmitting the attribute information output by the attribute information output means via the communication line,

An image identification device comprising:

27. The image identification device according to claim 25, wherein

The image receiving means,

E-mail receiving means for receiving an e-mail from the communication line power and extracting the input image information included in the e-mail;

An image identification device comprising:

28. In the image identification device according to claim 26,

The image transmitting means,

An electronic mail including the attribute information output by the attribute information output unit is transmitted to the communication line. Email sending means to send via

An image identification device comprising:

29. The image identification device according to claim 25, 26, 27 or 28, wherein the communication line is the Internet.

30. The image identification device according to claim 25, wherein

The communication line is a mobile communication line,

The image receiving means,

Mobile terminal receiving means for receiving the input image information from digital camera means via the mobile communication line;

An image identification device comprising:

3 1. The image identification device according to claim 26,

The communication line is a mobile communication line that performs communication by a mobile communication terminal, and the image transmission unit includes:

A mobile terminal transmitting means for transmitting the attribute information output by the attribute information output means to the other mobile communication terminal via the mobile communication line;

An image identification device comprising:

3 2. The image identification device according to claim 30 or 31,

The mobile communication line is a mobile phone communication line.