JP4754986B2 - Image browser system - Google Patents

Description

  The present invention relates to an image browser system that efficiently extracts and displays image data including a desired feature amount from a plurality of image data.

  A conventional image browser system enables browsing by selecting two axes from a multi-dimensional feature amount space and arranging reduced images (thumbnail) at corresponding coordinates (see, for example, Patent Document 1). FIG. 5 is a schematic diagram showing a graphic user interface in a conventional apparatus. In FIG. 5, images corresponding to the X and Y axes are displayed in the browse area 102 represented by the two axes selected by the X axis feature quantity selection button 111 and the Y axis feature quantity selection button 110. Is displayed.

  In the browse area 102 displayed in this way, the display scale can be changed by the enlargement / reduction ratio 101, and it is possible to browse more densely the locations where the thumbnails are dense. In addition, using three feature quantities, browsing in a three-dimensional space is also possible.

JP-A-11-175535 (first page, FIG. 1)

  However, the prior art has the following problems. The conventional image browser system handles only two feature quantities in the two-dimensional space display, and it is difficult to find the relationship with more feature quantities even if the scale change function is used together. .

  In the system configuration in which the three feature quantities are displayed in a three-dimensional space, since the three-dimensional space is projected onto a display that is a two-dimensional space, a three-dimensional overlap occurs, and the browsing prospect is poor. Furthermore, since it is necessary to rotate the three-dimensional space in order to grasp the whole, there is a problem that the processing amount increases and the usability is not good.

  The present invention has been made to solve the above-described problems, and reduces the amount of calculation required for display processing during image browsing, and allows an image to easily visually recognize the relationship with more feature values. The purpose is to obtain a browser system.

An image browser system according to the present invention includes: an image feature amount storage unit that stores an image feature amount table in which image data and feature amount information related to a plurality of feature amounts included in the image data are associated; and input image data A feature amount extracting unit that extracts a plurality of feature amounts to generate feature amount information, associates input image data with the feature amount information, and updates an image feature amount table stored in the image feature amount storage unit; Two-dimensional map feature quantity setting means for setting the first feature quantity and the second feature quantity for the two-dimensional map, and the first feature quantity from the image feature quantity table stored in the image feature quantity storage unit Or search means for searching for image data including at least one of the second feature quantity and generating a thumbnail list in which feature quantity information corresponding to the image data is associated; and the first feature quantity And a two-dimensional map having the second feature quantity as two axes, a display position of image data on the two-dimensional map is specified from the first feature quantity and the second feature quantity, and the display means An image browser system comprising a layout constructing means for displaying, wherein a reference image is set from image data displayed in a two-dimensional map based on the displayed two-dimensional map and associated with the reference image And an additional condition setting means for setting the third feature value, wherein the layout construction means includes the thumbnail list searched by the search means, the reference image and the third feature value set by the additional condition setting means. based on the third characteristic quantity included in the reference image, the similarity calculated from the third feature amount included in the image data other than the reference image, is displayed on the display unit 2 Among the image data other than the reference image on the original map is to reconstruct the two-dimensional map for highlighting the high image data similarity between the reference image with respect to the third feature amount.

  According to the present invention, a reference image serving as a reference is designated from the two-dimensional map, a third feature value associated with the reference image is designated, and the third feature value included in the reference image is used as a basis. By obtaining the similarity, image data having high similarity to the reference image can be highlighted on the two-dimensional map for the third feature amount, and the amount of calculation required for display processing can be reduced during image browsing. Thus, it is possible to obtain an image browser system capable of visually recognizing the relation with more feature amounts.

Hereinafter, a preferred embodiment of an image browser system of the present invention will be described with reference to the drawings.
The image browser system of the present invention selects a reference image from image data displayed on a two-dimensional map, and designates a new third feature amount associated with the reference image, thereby providing a third feature amount. The other image data having a high similarity to the reference image is identified and highlighted on the two-dimensional map.

Embodiment 1 FIG.
FIG. 1 is a configuration diagram of an image browser system according to Embodiment 1 of the present invention. The image browser system 10 in FIG. 1 includes an image feature amount storage unit 11, a feature amount extraction unit 12, a two-dimensional map feature amount setting unit 13, a search unit 14, a layout construction unit 15, and an additional condition setting unit 16. Is done. Furthermore, the display means 20 which displays a two-dimensional map outside is provided.

  The image feature amount storage unit 11 is configured to associate feature amount information quantitatively indicating how much a plurality of predetermined feature amounts are included in each input image data with each input image data. It is a memory | storage part which memorize | stores an image feature-value table.

  The feature amount extraction unit 12 quantitatively extracts a plurality of predetermined feature amounts from input image data captured from the outside, generates feature amount information corresponding to the captured input image data, and captures the feature amount information. The image feature quantity table stored in the image feature quantity storage unit 11 is updated in association with the input data and the generated feature quantity information.

  When the user wants to extract a desired image from the input image data stored in the image feature amount storage unit 11, the two-dimensional map feature amount setting unit 13 includes two feature amounts included in the desired image. Setting means for designating (corresponding to the first feature value and the second feature value). In this way, images having high relevance to these feature quantities are extracted using the two feature quantities as keys.

  The search means 14 converts the information about the two feature quantities designated by the feature quantity setting means 13 for the two-dimensional map into a query in order to search for the feature quantities of the X-axis and the Y-axis for the two-dimensional map, respectively. The image feature amount storage unit 11 is accessed. By this search, the search means 14 can extract input image data and feature amount information including at least one of the two specified feature amounts from the image feature amount table.

  Further, the search unit 14 generates a thumbnail (reduced image) necessary for display based on the extracted input image data and feature amount information, and associates the generated thumbnail with the X-axis and Y-axis feature amounts. Generate a thumbnail list for display. In the present invention, in order to cope with a case where a new third feature amount is specified later, feature amount information relating to all feature amounts other than the X-axis and the Y-axis is also attached to this thumbnail list. Has been.

  The layout construction unit 15 acquires the thumbnail list generated by the search unit 14, generates a two-dimensional map related to the two specified feature amounts based on the thumbnail list, and outputs the two-dimensional map to the display unit 20.

  The additional condition setting means 16 is a setting means for the user who has viewed the two-dimensional map displayed on the display means 20 to add further search conditions based on the displayed thumbnails. More specifically, the user selects one desired thumbnail from the displayed thumbnails and sets a reference thumbnail (corresponding to a reference reduced image). Further, the user sets the third feature amount as a feature amount other than the first feature amount and the second feature amount that have already been specified among the feature amounts included in the reference thumbnail.

  On the other hand, the layout construction unit 15 is based on the reference thumbnail and the third feature amount set by the additional condition setting unit 16 and the feature amount information included in the thumbnail list acquired from the search unit 14. Then, a comparison operation is performed between the third feature amount included in the reference thumbnail and the third feature amount included in the other thumbnails.

  Then, the layout construction unit 15 specifies one or more thumbnails having a high similarity to the reference thumbnail for the third feature amount based on the calculation result, and highlights the specified thumbnail in a two-dimensional manner. The map is reconstructed and displayed on the display means 20. As a result, the user can identify a thumbnail having a high degree of similarity of the third feature amount based on the highlighted content with respect to the reference thumbnail selected from the two-dimensional map as the image that the user wants to extract most. it can.

  As described above, according to the first embodiment, the corresponding thumbnails are displayed on the two-dimensional map based on the feature amounts for the two axes specified first, and then the reference thumbnail and the reference thumbnail and By specifying the third feature amount, it is possible to highlight a thumbnail having high similarity based on the third feature amount included in the reference thumbnail.

  As a result, in addition to the first feature amount and the second feature amount associated as a two-dimensional map, the similarity of the third feature amount can be added as an evaluation scale, and the feature amounts for the three axes can be calculated. It is possible to build an image browser system that can easily grasp the relationship. Further, when displaying information for three axes, there is an advantage that a projection process from a three-dimensional space to a two-dimensional space is unnecessary and a calculation amount required for the display process is small.

Embodiment 2. FIG.
In the second embodiment, a first specific example of highlighting by reconstructing a two-dimensional map by the layout construction unit 15 will be described.
FIG. 2 is a diagram showing an example of highlighting in Embodiment 2 of the present invention. In FIG. 2, the layout construction means 15 changes the size of the thumbnail to be highlighted with time.

  For example, the size of the thumbnail to be highlighted is x = x1 and y = y1 during time t1, and x = x2 and y = y2 during time t2. Then, the layout construction unit 15 repeatedly performs such highlighting processing and causes the display unit 20 to display the reconstructed two-dimensional map, thereby changing the size of the emphasized thumbnail temporally. it can.

  As a result, the user can visually and clearly confirm a thumbnail that is highly relevant to the designated reference thumbnail for the newly designated third feature amount. Note that the thumbnail for such highlighting is not limited to one thumbnail having the highest similarity to the reference thumbnail. It is also possible to sequentially highlight a plurality of thumbnails in descending order of similarity.

  As described above, according to the second embodiment, for the newly designated third feature amount, a thumbnail having a high similarity to the designated reference thumbnail can be displayed with the size changed at regular intervals. . Thereby, even when the similarity of the third feature amount is taken as an evaluation scale, the feature amount information for the three axes can be easily visually confirmed on the two-dimensional map. Further, when displaying information for three axes, there is an advantage that a projection process from a three-dimensional space to a two-dimensional space is unnecessary and a calculation amount required for the display process is small.

Embodiment 3 FIG.
In the third embodiment, a second specific example of highlighting by reconstructing a two-dimensional map by the layout constructing means 15 will be described.
FIG. 3 is a diagram showing an example of highlighting in the third embodiment of the present invention. In FIG. 3, a thumbnail (ref) is a reference thumbnail designated by the additional condition setting means 16. The thumbnail (S1) is the thumbnail having the highest similarity with the reference thumbnail for the third feature amount among the thumbnails displayed on the two-dimensional map. Further, the thumbnail (S2) is the thumbnail having the third feature amount having the second highest similarity to the reference thumbnail after the thumbnail (S1).

  In the third embodiment, the reference thumbnail (ref) and the thumbnails (S1) and (S2) having a high degree of similarity with respect to the third feature amount are displayed by connecting with a line to perform highlighting. More specifically, the layout construction unit 15 includes each of the third feature amounts included in all the thumbnails displayed on the two-dimensional map, and the third feature amount included in the reference thumbnail (ref). Comparison with the feature amount is performed, and thumbnails (S1) and (S2) are specified as thumbnails with high similarity.

  Furthermore, the layout construction means 15 displays a line L1 for connecting the reference thumbnail (ref) and the thumbnail (S1) and a line L2 for connecting the thumbnail (S1) and the thumbnail (S2). The coordinates of the start and end points of L1 and L2 are specified and displayed on the display means 20.

  As shown in FIG. 3, the designated reference thumbnail (ref) and the thumbnail (S1) having the highest similarity with the designated third feature amount are first connected by a line L1. It is. Further, the thumbnail (S2) having the next highest similarity is connected to the thumbnail (S1) by a line L2. Thereby, in a state where the two-dimensional distribution is displayed, it is possible to display a thumbnail that is highly relevant to the third feature amount with respect to the reference thumbnail so that it can be easily visually confirmed.

  As described above, according to the third embodiment, it is possible to display a thumbnail having a high similarity to the designated reference thumbnail for the newly designated third feature amount by connecting the reference thumbnail with a line. Thereby, even when the similarity of the third feature amount is taken as an evaluation scale, the feature amount information for the three axes can be easily visually confirmed on the two-dimensional map. Further, when displaying information for three axes, there is an advantage that a projection process from a three-dimensional space to a two-dimensional space is unnecessary and a calculation amount required for the display process is small.

Embodiment 4 FIG.
In the fourth embodiment, a third specific example of highlighting by reconstructing a two-dimensional map by the layout construction unit 15 will be described.
FIG. 4 is a diagram showing an example of highlighting in the fourth embodiment of the present invention. In FIG. 4, the layout construction means 15 changes the display position of the thumbnail to be highlighted temporally.

  For example, the position of the thumbnail to be highlighted is the coordinate position indicated by (x1, y1) during the time t1, and is moved to the position indicated by (x2, y2) during the time t2. Then, the layout construction unit 15 can repeatedly change the position of the emphasized thumbnail by repeatedly performing such highlight display processing and displaying the reconstructed two-dimensional map on the display unit 20. .

  As a result, the user can visually and clearly confirm a thumbnail that is highly relevant to the designated reference thumbnail for the newly designated third feature amount. Note that the thumbnail for such highlighting is not limited to one thumbnail having the highest similarity to the reference thumbnail. It is also possible to sequentially highlight a plurality of thumbnails in descending order of similarity.

  As described above, according to the fourth embodiment, for the newly designated third feature amount, a thumbnail having a high similarity to the designated reference thumbnail can be displayed with its position changed at regular intervals. . Thereby, even when the similarity of the third feature amount is taken as an evaluation scale, the feature amount information for the three axes can be easily visually confirmed on the two-dimensional map. Further, when displaying information for three axes, there is an advantage that a projection process from a three-dimensional space to a two-dimensional space is unnecessary and a calculation amount required for the display process is small.

  In the above-described embodiment, the case where one third feature amount is added as a new evaluation scale has been described, but the present invention is not limited to this. It is also possible to specify a plurality of feature quantities as a new evaluation scale. For example, when the fourth feature value is used as an evaluation scale in addition to the third feature value, a new similarity calculated from the similarity of the third feature value and the similarity of the fourth feature value By using a degree (for example, an average value of both similarities or a mean square value), the same highlighting can be performed. As a result, it is possible to easily visually display information on feature amounts for four axes or more on a two-dimensional map.

  In the present invention, thumbnails having the same feature values for the two specified axes are displayed at the same position. Therefore, when there are a plurality of thumbnails having the same feature amount for the two specified axes, the layout construction unit 15 slightly changes one or both of the X-axis and Y-axis coordinates of the thumbnail to be displayed. By doing so, it is displayed on the display means 20 while avoiding duplication. As a result, it is possible to display a plurality of thumbnails having the same value without being completely hidden while overlapping.

  In the case of a two-dimensional map display as in the present invention, a correlated color temperature based on image data can be used as a feature quantity to be extracted by the feature quantity extraction unit 12. In this case, since the physical quantity (numerical value) of the color temperature matches the visual sensitivity, display with improved human browsing efficiency is possible.

  In the above-described embodiment, the case where the retrieval unit 14 generates a reduced image has been described. However, the present invention is not limited to this. For example, if the size of the image data is large enough to be displayed on a two-dimensional map, it is not always necessary to use a reduced image, and the same effect can be obtained even if image data that has not been reduced is used. .

It is a block diagram of the image browser system in Embodiment 1 of this invention. It is a figure which shows an example of the highlight display in Embodiment 2 of this invention. It is a figure which shows an example of the highlight display in Embodiment 3 of this invention. It is a figure which shows an example of the highlight display in Embodiment 4 of this invention. It is the schematic diagram which showed the graphic user interface in a conventional apparatus.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Image browser system, 11 Image feature-value memory | storage part, 12 Feature-value extraction means, 13 2-dimensional map feature-value setting means, 14 Search means, 15 Layout construction means, 16 Additional condition setting means, 20 Display means

Claims (6)

An image feature quantity storage unit that stores an image feature quantity table in which image data and feature quantity information relating to a plurality of feature quantities included in the image data are associated;
A plurality of feature amounts are extracted from the input image data to generate feature amount information, the image feature amount stored in the image feature amount storage unit in association with the input image data and the feature amount information. Feature quantity extraction means for updating the table;
Two-dimensional map feature quantity setting means for setting a first feature quantity and a second feature quantity for a two-dimensional map;
Search for image data including at least one of the first feature amount and the second feature amount from the image feature amount table stored in the image feature amount storage unit, and correspond to the image data Search means for generating a thumbnail list in which feature quantity information to be associated is associated;
A two-dimensional map is constructed with the first feature quantity and the second feature quantity as two axes, and the display position of the image data on the two-dimensional map from the first feature quantity and the second feature quantity An image browser system comprising: a layout construction means for identifying and displaying on a display means,
Based on the displayed two-dimensional map, an additional condition setting that sets a reference image from the image data displayed in the two-dimensional map and sets a third feature amount associated with the reference image Further comprising means,
The layout construction unit is included in the reference image based on the thumbnail list searched by the search unit and the reference image and the third feature amount set by the additional condition setting unit. An image other than the reference image on the two-dimensional map displayed on the display means by obtaining a similarity between the third feature amount and a third feature amount included in image data other than the reference image An image browser system for reconstructing a two-dimensional map for highlighting image data having a high similarity to the reference image in the data with respect to the third feature amount . The image browser system according to claim 1,
The layout construction means, when there is image data in which the first feature quantity and the second feature quantity have the same value and display positions overlap in the constructed two-dimensional map, And an image browser system, wherein at least one of the second feature values is changed to reconstruct a two-dimensional map to prevent duplication of display positions. The image browser system according to claim 1 or 2,
The layout constructing means constructs a two-dimensional map so that a predetermined number of image data is displayed in order of increasing similarity so that the display size is changed temporally. The image browser system according to claim 1 or 2,
The layout construction means constructs a two-dimensional map so that a predetermined number of image data are connected and displayed in descending order of similarity. The image browser system according to claim 1 or 2,
The layout constructing means constructs a two-dimensional map so that a predetermined number of image data is displayed in order of increasing similarity so that a display position is changed temporally. The image browser system according to any one of claims 1 to 5,
The image browser system, wherein the feature amount extraction unit extracts the plurality of feature amounts using a correlated color space based on the input image data.

Images