KR20110068587A - Content based retrieval service method for images with location information - Google Patents

Abstract

PURPOSE: A content based retrieval service method for images with location information is provided to easily obtain search results on the basis of an accurate location by using a specific image and a location information included image. CONSTITUTION: A map screen is displayed on an initial screen. If a user sets an initial location on the map screen, a marker is marked on the location. A user inputs an image or text keyword into a search dialog box presented on the map screen. The keyword type is checked. A similar image is retrieved from an image database by handling the process according to the keyword type, and is then displayed to the user.

Description

CONTENT BASED RETRIEVAL SERVICE METHOD FOR IMAGES WITH LOCATION INFORMATION}

The present invention relates to a content-based search service method for an image including location information. The present invention relates to an image including location information that allows a user to extract latitude / longitude information of a desired location using a specific image as information. The present invention relates to a content-based search service method.

With the spread of the Internet, a lot of video contents have been generated and shared on the web in recent years.

Such content is so vast that users use search to find the content they want.

However, due to space and time constraints, the range of content that a person can access is limited. For this reason, an efficient information retrieval method has been studied, but at present, it is not far from the text-based retrieval method in the actual application stage.

Video content visually contains information about a specific place.

People rely on visual information when they make appointments or travel itineraries, and are searching for video content to get more information. Currently, information is performed by converting the visible information into text.

For example, if someone sees a picture of "New York Confectionery in front of Gangnam Station" on someone's blog, the user can collect location information through the text "Gangnam Station, New York Confectionery".

However, if a photo posted on a blog contains location information and can be searched with the content of the picture itself, at least the location information for the same picture can be obtained more effectively.

FIG. 33 is a diagram illustrating an image displayed on a map as a conventional search result, and FIG. 34 is a diagram illustrating an image represented by a list as a conventional search result.

Meanwhile, examples of the conventional image search service are as shown in FIGS. 33 and 34.

In detail, FIG. 33 is a diagram showing an image related to a query on the screen when a query is input as text to a map service. FIG. 34 shows results obtained by inputting text as a query through a general web search service. This is the figure displayed on the screen.

Conventional image retrieval service provides a search result to a user through a method of determining whether it is related to a query by using a file name of a picture or a title or content of a post posted with the picture.

In FIG. 33, image information may be extracted from a map. However, when a user first performs a text search, the search engine displays a link related to a query input by the user on a map. Thereafter, the user must perform a secondary procedure of following a link through an operation of zooming / moving / moving the map.

In addition, in the case of FIG. 34, the similarity degree to the query is determined in the same manner as in FIG. 33. In this case, there is a limit in reflecting the mood or color of the picture actually desired by the user.

The present invention has been made to solve the above problems, and includes a location information that allows a user to easily obtain a precise location-based search results using a specific image and an image including location information stored in a database. Its purpose is to provide a content-based retrieval service method for video.

Another object of the present invention is to extract a feature owned by the image itself, select an image having similar features, and provide meaningful information. To provide a search service method.

It is still another object of the present invention to provide a map screen through a user interface, while improving the user's search convenience, and content-based search for an image including location information for sorting the search results according to the content similarity of the image. To provide a service method.

In order to achieve the above object, a content-based search service method for an image including location information according to the present invention includes a map screen display step of displaying a map screen on an initial screen, and a user initializing the map screen. When the location is set, an initial location selection step of displaying a marker at a corresponding location, a search word input step in which a user inputs an image or text search word in a search dialog box displayed on the map screen, and a search word input in the search dialog box And a similar image search step of searching for a similar image in an image database and displaying the same to a user by processing a process according to the type of the search word.

In addition, in the initial location selection step, the user may set an initial location by inputting a keyword on the map screen.

In addition, in the initial position selection step, the user may set an initial position by performing a zoom in, zoom out, and move function on the map screen.

The initial position selection step may set a search radius around the initial position set by the user.

Also, in the search word input step, the search dialog may include a radio button for distinguishing a search using an image or a search using text.

In addition, in the search word input step, when the image is input to the search dialog box for searching, the file format of the image may be a general format such as jpg, bmp, or png, and the image may include location information.

In addition, in the similar image display step, when the type of the search word is an image, a feature of the corresponding image is extracted, an identifier is generated, compared with an identifier list of the image database, and a similar image list is obtained to the user. Can be displayed.

In addition, the similar image displaying step may include a size normalization process of converting the input image into a W (width) × H (height) size to form a thumbnail image of the input image, and converting the input image into W × H. A partial block dividing step of dividing into four partial blocks, a pixel value setting step of setting an average pixel value of the partial block as a pixel value of the thumbnail image, and extracting image characteristics of the thumbnail image It may include an identifier generating step for generating a thumbnail identifier.

In addition, the similar image display step may include a size normalization step of constantly converting the size of the input image, a concentric circle mapping process of dividing the normalized input image into a unit of radius, and dividing the normalized input image into units of an angle again; An average pixel distribution calculation step of calculating an average pixel distribution of all regions in the ring partitioned within the concentric circle, and an absolute value distribution of a difference value between the average pixel value obtained in the average pixel distribution calculation step and the average pixel value of the neighboring ring An average pixel secondary variation that calculates an absolute value distribution of a difference value between a primary variation obtained by the average pixel primary variation distribution calculation step and a primary variation value obtained by the average pixel primary variation distribution calculation process The process of calculating the distribution and the mean for one region in the ring and the region adjacent to the ring and the region located opposite the ring A rotational disparity calculation step of calculating a difference between pixel distributions and a feature distribution converting the feature distributions of the average pixel distribution, the primary variation distribution, the secondary variation distribution, and the rotational variation distribution into a bit string using a hash function. And a hashing step and an identifier generation step of generating a still picture descriptor from the bit string.

In addition, the similar image display step includes an input image segmentation process of dividing an input image into N × M partial blocks, and 0 °, 45 °, 90 °, 135 °, 180 °, and 225 as feature values for the partial block. An eight-way edge feature value extraction process for extracting features for eight direction edge components of °, 270 °, and 315 °, and a representative edge component value of the partial block are determined and the values are stored as identifiers. It may include an identifier generation process.

The similar image display step may include a histogram generating step of extracting a histogram of a three-dimensional color space of an input image, a histogram quantization step of quantizing the number of ranks of the histogram into L1, L2, and L3, and the frequency of the histogram. A histogram normalization step of normalizing the value to have a value of 0 to 255, and an identifier generation step of generating an identifier by storing the frequency value of the normalized histogram in a single byte space.

The method may further include a similar image re-searching step of re-entering the similar image searching step by re-entering the searched similar image into the search dialog box after the similar image searching step.

The method may further include registering an image in which a user selects a specific location on the map screen and uploads an image including the location information.

The image registration step may include a specific location selection process of selecting a specific location on the map screen, a latitude / longitude extraction process of extracting latitude / longitude values of the specific location, and an image including location information of the user. It may include a video upload process for uploading a location value registration process for registering a location value from the metadata of the uploaded image.

In addition, when the type of the search word is text, the similar image displaying step may be compared with the image database to obtain an image list having similar contents and display the same to the user.

As described above, the content-based search service method for an image including location information according to the present invention enables a user to easily obtain an accurate location-based search result using a specific image and an image including location information stored in a database. It can work.

In addition, by extracting a feature owned by the image itself, it is effective to increase the user's search efficiency by selecting a similar image and providing meaningful information.

In addition, by providing a map screen as a user interface, it is possible to sort the search results according to the similarity of the contents of the image while increasing the user's search convenience.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. First, it should be noted that the same components or parts in the drawings represent the same reference numerals as much as possible. In describing the present invention, detailed descriptions of related well-known functions or configurations are omitted in order not to obscure the gist of the present invention.

1 to 4 are block diagrams of a content-based search service method for an image including location information according to an embodiment of the present invention, and FIG. 5 includes location information according to an embodiment of the present invention. A diagram showing the overall algorithm of a content-based retrieval service method for video.

Content-based search service method for the image including the location information according to an embodiment of the present invention, as shown in Figure 1, the map screen display step (S100), the initial location selection step (S200), A search word input step S300, a search word type checking step S400, and a similar image search step S500 are included.

The map screen display step (S100) is a step of displaying a map screen as an initial screen for searching, as shown in FIGS. 1 and 5.

6 is a diagram illustrating a state of setting an initial position on a map screen, FIG. 7 is a diagram illustrating a state of setting a search radius, and FIG. 28 is a diagram illustrating an initial position setting and a search word input.

The initial location selection step (S200) is a step of displaying a marker at a corresponding location when the user sets an initial location on the map screen.

In the initial position selection step (S200), as shown in FIG. 6, the user inputs a keyword on the map screen to set an initial position, or performs a zoom in, zoom out, and move function on the map screen to set the initial position. By setting, the search condition can be set.

In addition, in the initial position selection step (S200), as shown in FIG. 7, the user may additionally set a search condition by setting a search radius around the initial position.

As described above, by setting the search radius, the user may select the image data having the latitude / longitude value within the initial position and the search radius.

The search word input step (S300) is a step in which a user inputs an image or text search word in a search dialog box displayed on the map screen.

As shown in FIG. 28, the search dialog is a window in which a user can input a search word, and includes a radio button for distinguishing a search using an image or a search using text. Therefore, it is possible to distinguish whether a search using an image or a search using text through the radio button.

On the other hand, in the case of searching by inputting an image in the search dialog box, the file format of the image may be a general format such as jpg, bmp, or png, and the image may include location information.

The search word type checking step (S400) is a step of checking whether a search word input in the search dialog box is an image or text.

The similar image retrieval step (S500) is a process of searching for a similar image in an image database by processing a process according to the type of the search word and displaying the same to the user.

In the similar image retrieval step (S500), when the search word, that is, the query is an image query, a process for the image query is performed, and when the query is a text query, the process is performed for the text query.

If the search term type is an image, the similar image search step (S500) extracts a feature of the corresponding image, generates an identifier, compares the identifier with an identifier list of an image database, and obtains an image list having similar contents to display to the user. do.

The image database may be configured to have an image file itself and to simultaneously store an identifier file corresponding to the image file through feature extraction and identifier generation for each image.

This is to easily obtain an image list having similar contents compared to the identifier list previously extracted after performing feature extraction and identifier generation for the query word when the image comes in as a query word.

FIG. 8 is a first block diagram of a similar image search step of a content-based search service method for an image including location information according to an embodiment of the present invention, and FIG. 9 is a diagram illustrating the algorithm of FIG. 8. FIG. 10 is a diagram illustrating thumbnail image generation, and FIG. 11 is a diagram illustrating a feature of storing a feature using a thumbnail image.

As shown in FIG. 8, the similar image retrieval step S500 includes a size normalization step S511, a partial block dividing step S512, a pixel value setting step S513, and an identifier generation step S514. can do.

The size normalization process (S511) is a process of converting the input image into a W (width) × H (height) size to form a thumbnail image of the input image.

Since the size normalization step S511 does not know the size of the input image to be entered in the query language in advance, it may be referred to as a normalization step for this point.

As shown in FIG. 10, the partial block dividing process S512 divides the input image into W × H partial blocks.

The pixel value setting step S513 is a step of setting an average pixel value for the partial block to a pixel value of the thumbnail image, as shown in FIG. 11.

The identifier generation step S514 is a process of generating a thumbnail identifier by extracting an image feature with respect to the thumbnail image.

As described above, the algorithm for generating the thumbnail image is shown in FIG. 9.

12 is a second block diagram of a similar image search step of a content-based search service method for an image including location information according to an embodiment of the present invention, and FIG. 13 is a diagram illustrating the algorithm of FIG. 14 is a diagram illustrating a zone setting through a concentric section mapping process, FIG. 15 is a diagram illustrating a pixel value variation distribution, and FIG. 16 is a diagram illustrating a rotational variation distribution. In addition, FIG. 17 is a diagram illustrating bit strings for four feature distributions, and FIG. 18 is a diagram illustrating bit string generation.

Meanwhile, as shown in FIG. 12, the similar image retrieval step S500 includes a size normalization step S521, a concentric division mapping step S522, an average pixel distribution calculation step S523, and an average pixel 1. The difference variation distribution calculation step (S524), the average pixel secondary variation distribution calculation step (S525), the rotational variation distribution calculation step (S526), the feature distribution hashing step (S527), and the identifier generation step (S528) It may include.

The size normalization process (S521) is a process of constantly converting the size of the input image.

As shown in FIG. 14, the concentric partition mapping process (S522) divides the normalized input image into a radius unit and divides the normalized input image into an angle unit and divides it into a predetermined region. In this case, the value of each zone may be determined as an average pixel value in the zone.

Meanwhile, the still image descriptor may be composed of four feature distributions.

Specifically, the feature distribution of the still image descriptor may be an average pixel distribution, a primary variation distribution of the average pixel, a secondary variation distribution of the average pixel, and a rotational variation distribution.

The average pixel distribution calculating step (S523) is a step of calculating an average pixel distribution of all regions in the ring partitioned in the concentric circles.

As shown in FIG. 15, the average pixel primary variation distribution calculation step S524 is used to calculate the absolute value distribution of the difference between the average pixel value obtained in the average pixel distribution calculation step S523 and the average pixel value of the neighboring ring. Calculation process.

The average pixel secondary variation distribution calculation step (S525) calculates an absolute value distribution of a difference value between the primary variation value obtained in the average pixel primary variation distribution calculation process (S524) and the primary variation value of a neighboring ring. It is a process.

As shown in FIG. 16, the rotational disparity distribution calculation process S526 calculates a difference between an average pixel distribution of one region in the ring, a neighboring region in the ring, and a region located opposite to the region in the ring. to be.

In the feature distribution hashing process (S527), as shown in FIGS. 17 and 18, a hash function is used for the feature distribution of the average pixel distribution, the primary variation distribution, the secondary variation distribution, and the rotational variation distribution. Is converted to a bit string.

The identifier generating step S528 is a step of generating a still image descriptor from the bit string.

As described above, the algorithm for generating the concentric partition-based identifier is as shown in FIG.

19 is a third block diagram of a similar image search step of a content-based search service method for an image including location information according to an embodiment of the present invention. FIG. 20 is a diagram illustrating an algorithm of FIG. 19. 21 illustrates edge-based identifier generation.

Meanwhile, as illustrated in FIG. 19, the similar image retrieval step S500 may include an input image segmentation step S531, an eight-way edge feature value extraction step S532, and an identifier generation step S533. .

The input image segmentation process S531 is a process of dividing an input image into N × M partial blocks.

As shown in FIG. 21, the eight-way edge feature value extracting process S532 may include 0 °, 45 °, 90 °, 135 °, 180 °, 225 °, as feature values for the N × M partial blocks. It is a process of extracting features for eight directions of edge components of 270 ° and 315 °.

The identifier generation step S533 determines a representative edge component value of each partial block and stores the value as an identifier.

As described above, an algorithm for generating an identifier using an edge feature is as shown in FIG. 20.

FIG. 22 is a fourth block diagram of a similar image search step of a content-based search service method for an image including location information according to an embodiment of the present invention, and FIG. 23 is a diagram illustrating an algorithm of FIG. 22. 24 is a diagram illustrating histogram extraction and quantization, and FIG. 25 is a diagram illustrating histogram normalization and storage.

Meanwhile, as shown in FIG. 22, the similar image retrieval step S500 includes a histogram generation step S541, a histogram quantization step S542, a histogram normalization step S543, and an identifier generation step S544. It may include.

As shown in FIG. 24, the histogram generating process S541 is a process of extracting a histogram of a three-dimensional color space of an input image.

The three-dimensional color space of the input image is RGB (Red, Green, Blue) or HSV (Hue, Saturation, Value) or YUV (Y: luminance signal, U: luminance signal and red component difference, V: luminance signal and blue Difference of components) space.

The histogram quantization step S542 is a step of quantizing the number of ranks of the histogram into L1, L2, and L3 for each of three dimensions.

The histogram normalization step S543 is a step of normalizing the frequency value of the histogram to have a value of 0 to 255, as shown in FIG.

The identifier generating step S544 is a step of generating an identifier by storing the normalized frequency value of the histogram in a 1-byte space.

As described above, the algorithm for generating the identifier using the color histogram is as shown in FIG.

Meanwhile, in the similar image searching step S500, a database corresponding to the selected location information may be searched.

Specifically, image data having the initial position selected by the user and latitude / longitude values within the search radius set by the user may be selected using metadata of the image itself, and if the user does not specifically select a radius, the full screen area Image data having latitude / longitude values can be selected from.

In the similar image retrieval step S500, an image having a high similarity may be selected using an image identifier.

In detail, the image having high similarity between the image identifier and the bitstream may be sorted as a result having a high priority among the data selected using the location information. This is because image matching speed using image identifier is relatively slow compared to search speed using metadata.

29 is a diagram illustrating a search result display.

In addition, in the similar image search step S500, as shown in FIG. 29, the search results may be displayed on a map screen. In this case, the results may be sorted and displayed according to a user's request.

The red icon on the map screen is a marker for location information of a search result, and the blue icon is a marker indicating an initial location set by the user for searching.

On the other hand, in the similar image search step (S500), as shown in FIG. 5, when the search word, that is, the query is a text query, a process for the text query may be performed.

In detail, when the type of the search word is text, an image list having similar contents may be obtained and displayed to the user after comparing with the image database.

Meanwhile, in the content-based search service method for an image including location information according to an embodiment of the present invention, as shown in FIG. And a search word input step S300, a search word type check step S400, a similar image search step S500, and a similar image rescan step S600.

The map screen display step (S100), the initial position selection step (S200), the search word input step (S300), the search word type checking step (S400), and the similar image search step (S500) are the same as described above.

30 is a diagram illustrating re-search using search results.

The similar image rescanning step (S600) is a step of re-entering the similar image searching step (S500) by re-inputting the found similar image into a search dialog box.

The user may use the image obtained as a result of the search again as a query word in the similar image re-search step S600, and as shown in FIG. It may be an image selected for.

In addition, the content-based search service method for the image including the location information according to an embodiment of the present invention, as shown in Figure 3, the map screen display step (S100) and the initial location selection step (S200) And a search word input step S300, a search word type checking step S400, a similar image search step S500, and an image registration step.

The map screen display step (S100), the initial position selection step (S200), the search word input step (S300), the search word type checking step (S400), and the similar image search step (S500) are the same as described above.

The image registration step (S700) is a step in which a user selects a specific location on the map screen and uploads an image including the location information.

By uploading the location information of the image in the image registration step (S700), the user can easily register the location information without knowing the actual latitude / longitude value on the map screen.

FIG. 26 is a block diagram of an image registration step of a content-based search service method for an image including location information according to an embodiment of the present invention, and FIG. 27 is a diagram illustrating an algorithm of FIG.

As illustrated in FIG. 26, the image registration step S700 includes a specific position selection process S710, a latitude / longitude extraction process S720, an image upload process S730, and a position value registration process S740. Include.

The specific location selection process (S710) is a process in which a user selects a specific location on the map screen.

The latitude / longitude extraction process S720 is a process of extracting the latitude / longitude value of the specific position.

The image upload process (S730) is a process in which a user uploads an image including location information.

The position value registration step S740 is a process of registering a position value from metadata of the uploaded image.

As described above, the algorithm for image registration by the user is as shown in FIG.

In addition, the content-based search service method for an image including location information according to an embodiment of the present invention, as shown in Figure 4, the map screen display step (S100) and the initial location selection step (S200) And a search word input step S300, a search word type checking step S400, a similar image search step S500, a similar image rescanning step S600, and an image registration step S700.

The map screen display step (S100), the initial position selection step (S200), the search word input step (S300), the search word type checking step (S400), the similar image search step (S500), the similar image re-search step ( S600) and the image registration step (S700) is the same as described above.

31 is a diagram illustrating a process for an image query, and FIG. 32 is a diagram illustrating a process for a text query.

In the content-based search service method for an image including location information according to an embodiment of the present invention, the process algorithm for the image query is illustrated in FIG. 31, and the process algorithm for the text query is illustrated in FIG. 32. As shown in.

As described above, the present invention goes beyond the existing one-dimensional keyword-centered image retrieval method and can automatically and effectively set the search conditions, and the link method between the user and the content in that the image obtained from the search result is meaningful information. New directions can be given.

Accordingly, the present invention relates to a video retrieval service and a location information providing service based on location information, a video retrieval service based on image content information, a photo album management field of a personal desktop and a social network based on location information. It can be applied to the field of video album and sharing service.

As described above with reference to the drawings illustrating a content-based search service method for an image including location information according to the present invention, the present invention is not limited to the embodiments and drawings disclosed herein. Of course, various modifications may be made by those skilled in the art within the technical scope of the present invention.

1 to 4 are block diagrams of a content-based search service method for an image including location information according to an embodiment of the present invention.

5 is a diagram illustrating an overall algorithm of a content-based search service method for an image including location information according to an embodiment of the present invention.

6 is a diagram illustrating a state of setting an initial position on a map screen, and FIG. 7 is a diagram illustrating a state of setting a search radius.

8 is a first block diagram of a similar image search step of a content-based search service method for an image including location information according to an embodiment of the present invention.

9 is a diagram illustrating an algorithm of FIG. 8, FIG. 10 is a diagram illustrating thumbnail image generation, and FIG. 11 is a diagram showing a state of storing a feature using a thumbnail image.

12 is a second block diagram of a similar image search step of a content-based search service method for an image including location information according to an embodiment of the present invention.

FIG. 13 is a diagram illustrating an algorithm of FIG. 12, FIG. 14 is a diagram illustrating zoning through a concentric partition mapping process, FIG. 15 is a diagram illustrating a pixel value variation distribution, and FIG. 16 is a diagram showing a rotational variation distribution. 17 shows a bit string for four feature distributions, and FIG. 18 shows bit string generation.

19 is a third block diagram of a similar image search step of a content-based search service method for an image including location information according to an embodiment of the present invention.

20 illustrates the algorithm of FIG. 19, and FIG. 21 illustrates edge-based identifier generation.

FIG. 22 is a fourth block diagram of a similar image search step of a content-based search service method for an image including location information according to an embodiment of the present invention. FIG.

FIG. 23 illustrates the algorithm of FIG. 22, FIG. 24 illustrates histogram extraction and quantization, and FIG. 25 illustrates histogram normalization and storage.

FIG. 26 is a block diagram of an image registration step of a content-based search service method for an image including location information according to an embodiment of the present invention. FIG.

FIG. 27 is a diagram illustrating the algorithm of FIG. 26, FIG. 28 is a diagram illustrating initial position setting and a search word input, FIG. 29 is a diagram showing a search result display, and FIG. 30 is a diagram showing re-search using search results.

31 illustrates a process for an image search term.

32 illustrates a process for text search terms.

FIG. 33 is a diagram showing an image displayed on a map as a conventional search result, and FIG. 34 is a diagram showing an image represented as a list as a conventional search result.

<Description of Symbols for Main Parts of Drawings>

S100: Map screen display step S200: Initial location selection step

S300: search word input step S400: search word type checking step

S500: Similar image search step S511, S521: Size normalization process

S512: Partial block dividing step S513: Pixel value setting step

S514, S528, S533, S544: Identifier generation process

S522: Concentric segment mapping process S523: Average pixel distribution calculation process

S524: Calculation of average pixel primary variation distribution

S525: average pixel secondary variation distribution calculation process

S526: Rotating Dispersion Distribution Calculation Process

S527: Characteristic distribution hashing process S531: Input image segmentation process

S532: Eight-way edge feature value extraction process

S541: Histogram Generation Process S542: Histogram Quantization Process

S543: Histogram normalization process S600: Similar image rescanning step

S700: Image registration step S710: Specific position selection process;

S720: Latitude / Longitude Extraction Process S730: Image Upload Process

S740: Position value registration process

Claims (15)

A map screen display step of displaying a map screen as an initial screen; An initial location selection step of displaying a marker at a corresponding location when a user sets an initial location on the map screen; A search word input step of a user inputting an image or text search word in a search dialog box displayed on the map screen; A search word type checking step of checking a type of a search word input in the search dialog box; And And a similar image searching step of processing a process according to the type of the search term to search for a similar image in an image database and displaying the same to a user. The method of claim 1, The initial position selection step, A content-based search service method for an image including location information, wherein a user sets an initial location by inputting a keyword on the map screen. The method of claim 1, The initial position selection step, A content-based search service method for an image including location information, wherein the user sets an initial location by performing a zoom in, zoom out, and movement function on the map screen. The method of claim 1, The initial position selection step, A content-based search service method for an image including location information, wherein a search radius is set based on the initial location set by a user. The method of claim 1, In the search word input step, The search dialog box is a content-based search service method for an image including the location information, characterized in that it comprises a radio button for distinguishing the search using the image or the search using the text. The method of claim 5, In the search word input step, When searching by inputting an image into the search dialog box, the file format of the image is a general format such as jpg, bmp, or png, and the image includes position information. Content-based retrieval service method. The method of claim 1, In the similar image display step, When the type of the search word is an image, location information of the feature is extracted, the identifier is generated and compared with the identifier list of the image database, and the image list having similar contents is obtained and displayed to the user. Content-based retrieval service method targeting images. The method of claim 7, wherein The similar image display step, A size normalization process of converting the input image into a W (width) × H (height) size to form a thumbnail image of the input image; A partial block dividing step of dividing the input image into W × H partial blocks; A pixel value setting step of setting an average pixel value for the partial block to a pixel value of the thumbnail image; And And an identifier generating step of generating a thumbnail identifier by extracting an image feature of the thumbnail image and generating a thumbnail identifier of the thumbnail image. The method of claim 7, wherein The similar image display step, A size normalization process of constantly converting the size of the input image; A concentric circle mapping process for dividing the normalized input image into radius units and dividing the normalized input image into angle units and dividing the normalized input image into regular regions; An average pixel distribution calculation step of calculating an average pixel distribution of all regions in the ring partitioned within the concentric circles; An average pixel primary variation distribution calculation step of calculating an absolute value distribution of a difference value between an average pixel value obtained in the average pixel distribution calculation step and an average pixel value of a neighboring ring; An average pixel secondary variation distribution calculation step of calculating an absolute value distribution of a difference value between the primary variation value obtained in the average pixel primary variation distribution calculation process and the primary variation value of a neighboring ring; A rotational disparity calculation step of calculating a difference between an average pixel distribution for one region in the ring, an adjacent region in the ring, and a region located opposite the ring; A feature distribution hashing step of converting the feature distribution of the average pixel distribution, the primary variation distribution, the secondary variation distribution, and the rotational variation distribution into a bit string using a hash function; And And an identifier generation step of generating a still image descriptor from the bit string. The method of claim 7, wherein The similar image display step, An input image segmentation process of dividing the input image into N × M partial blocks; An eight-way edge feature that extracts features for eight direction edge components of 0 °, 45 °, 90 °, 135 °, 180 °, 225 °, 270 °, and 315 ° as feature values for the partial block. Value extraction process; And And an identifier generating step of determining a representative edge component value of the partial block and storing the value as an identifier. The method of claim 7, wherein The similar image display step, A histogram generating step of extracting a histogram of a three-dimensional color space of an input image; A histogram quantization step of quantizing the number of ranks of the histogram into L1, L2, and L3; A histogram normalization step of normalizing the frequency value of the histogram to have a value of 0 to 255; And And an identifier generating step of generating an identifier by storing the normalized frequency value of the histogram in a one-byte space. The method of claim 1, After the similar image search step, And re-entering the similar image retrieval step by re-entering the retrieved similar image in the search dialog box. The method of claim 1, And a video registration step in which a user selects a specific location on the map screen and uploads an image including the location information. The method of claim 13, The image registration step, A specific location selection process of the user selecting a specific location on the map screen; A latitude / longitude extraction process of extracting a latitude / longitude value of the specific position; An image upload process of uploading an image including location information by a user; And And a location value registration process of registering a location value from metadata of the uploaded image. The method of claim 1, The similar video display step, If the type of the search word is a text, the content-based search service method for an image including the location information, characterized in that the image list is obtained and displayed to the user after comparing with the image database.

Images