Electronic device, keyword generation method, and program

Abstract

To generate a keyword based on an object included in an image even when it is difficult to set a keyword for retrieval such as an image.SOLUTION: A CPU 10 performs a predetermined image recognizing process on an image included in content, calculates a feature amount of an object displayed as the image, specifies a category to which the object belongs and a component part constituting the object on the basis of the feature amount, acquires a name (corresponding to a keyword) of the component part on the basis of the specified category and component part, and displays a name of the component part in association with the component part of the image. Further, the CPU 10 performs a dictionary retrieval using the name (corresponding to the keyword) of the component part displayed in association with the component part of the image, and displays a retrieval result.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an electronic device, a keyword generating method, and a program. [Background technology]

[0002] Conventionally, electronic dictionaries are known that incorporate multiple dictionary contents (hereinafter, simply referred to as contents), such as a Japanese dictionary, a Chinese-Japanese dictionary, an English-Japanese dictionary, a Japanese-English dictionary, and an encyclopedia. In electronic dictionaries, it is possible to search for information related to a word to be searched for (such as meaning, images, and translations into other languages) from multiple contents. In addition, on the Internet, it is possible to search for the meaning of a word, images and videos corresponding to the word, and related news and articles (WEB dictionary) by using the World Wide Web (WWW). For example, as an electronic dictionary, a plant recognition system that identifies the type of plant in an image based on the feature amounts of flowers and leaves contained in the image has been disclosed (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2002-203242 A Summary of the Invention [Problem to be solved by the invention]

[0004] As mentioned above, there are technologies that recognize objects in an image based on features contained in the image and display and output their names, descriptions, etc. However, there are no technologies that anticipate searching using the recognition results as keywords.

[0005] Therefore, an object of the present invention is to enable the generation of keywords based on objects contained in an image, even for objects such as images for which it is difficult to set keywords for searching. [Means for solving the problem]

[0006] The electronic device of the present invention is characterized by comprising an input unit for inputting an image, and a control unit for acquiring features of the input image and generating keywords to be used as search targets that are related to objects contained in the input image based on the acquired features. Effect of the Invention

[0007] According to the present invention, even for an object such as an image for which it is difficult to set keywords for a search, a keyword based on an object contained in the image can be generated. [Brief description of the drawings]

[0008] [Figure 1] 1 is a block diagram showing a configuration of an electronic device (electronic dictionary) 1 according to a first embodiment of the present invention. [Diagram 2] FIG. 1 is a perspective view showing the appearance of an electronic device 1 according to a first embodiment. [Diagram 3] 5 is a flowchart illustrating the operation of the electronic device 1 according to the first embodiment. [Figure 4] 5 is a flowchart illustrating the operation of the electronic device 1 according to the first embodiment. [Diagram 5] FIG. 2 is a schematic diagram showing an example of a display screen (home screen) of the electronic device 1 according to the first embodiment. [Figure 6] FIG. 2 is a schematic diagram showing an example of a display screen (search result screen) in the electronic device 1 according to the first embodiment. [Figure 7] FIG. 2 is a schematic diagram showing an example of a display screen (image) in the electronic device 1 according to the first embodiment. [Figure 8] 2 is a conceptual diagram showing an example of a feature amount list extracted from an image by image recognition processing by the electronic device 1 of the first embodiment. FIG. [Figure 9] 2 is a schematic diagram showing an example of a display screen (image) of an image recognition result by the electronic device 1 of the first embodiment. FIG. [Figure 10]FIG. 2 is a schematic diagram showing an example of a display screen (image) of a search result obtained by further searching from the display screen (image) of the image recognition result by the electronic device 1 of the first embodiment. [Figure 11] FIG. 4 is a conceptual diagram showing a system configuration of a second embodiment of the present invention. [Figure 12] FIG. 11 is a schematic diagram showing an example of a display screen (search result screen) on an electronic device (smartphone, personal computer) 82 according to the second embodiment. [Figure 13] FIG. 11 is a schematic diagram showing an example of a display screen (image) of electronic device 82 according to the second embodiment. [Figure 14] FIG. 11 is a schematic diagram showing an example of a display screen (image) of an image recognition result by electronic device 82 of the second embodiment. [Figure 15] FIG. 11 is a schematic diagram showing an example of a display screen (image) of a search result obtained by further searching from a display screen (image) of an image recognition result by electronic device 82 of the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0010] A. Configuration of the First Embodiment 1 is a block diagram showing the configuration of an electronic device (electronic dictionary) according to a first embodiment of the present invention. In the figure, electronic device (electronic dictionary) 1 includes a CPU 10, a key input unit 11, a display unit 12, a touch panel 13, and a memory 14.

[0011] The CPU 10 executes processing based on a predetermined program in response to user instructions input from the key input unit 11, transfers instructions and data to each functional unit, and generally controls the electronic device 1. In particular, the CPU 10 performs a predetermined image recognition process on an image included in the content, calculates (acquires) a feature amount of an object displayed as an image, generates a keyword related to the object of the image based on the feature amount, displays the generated keyword in association with the image, and further performs a dictionary search using the generated keyword and displays the search result.

[0012] More specifically, in the first embodiment, CPU 10 performs a predetermined image recognition process on an image included in the content, calculates the feature amount of an object displayed as an image, identifies the category to which the object belongs and the constituent parts that make up the object based on the feature amount, obtains the names of the constituent parts (corresponding to keywords) based on the identified category and constituent parts, and displays the names of the constituent parts in association with the constituent parts of the image. Furthermore, CPU 10 further performs a dictionary search using the names of the constituent parts (corresponding to keywords) displayed in association with the constituent parts of the image, and displays the search results.

[0013] Here, the feature amount is a numerical value of the shape, color, contrast, etc. of the object itself, or the shape, relative positional relationship, color, contrast, etc. of the components that make up the object. Also, the categories include, for example, vehicles (by vehicle type), ships (by ship type), buildings, natural objects (mountains, rivers, oceans), plants (by type), animals, people, etc.

[0014] The key input unit 11 is made up of a plurality of keys, and supplies key input to the CPU 10 in response to user operations. The display unit 12 is a device for displaying various screens under the control of the CPU 10, and is realized by, for example, a color LCD (Liquid Crystal Display) or an organic ELD (Electro-Luminescence Display). The touch panel 13 is a transmissive touch sensor provided on the screen of the display unit 12, and a touch position corresponds to a position (coordinate) on the screen, and it is possible to perform instruction operations on the screen, such as clicking and selecting icons, tabs, etc. displayed on the display unit 12, dragging and moving icons, and moving a cursor to a desired position on the screen, by touch operations using a user's finger, a stylus pen, etc.

[0015] The memory 14 stores an electronic dictionary program 141, an image analysis program 142, contents 143, and a keyword database (database) 144. The contents 143 and the keyword database 144 may be provided by separate storage media, and it is preferable that the images included in the contents 143 are photographed using an imaging device such as a digital camera or captured using an image input device such as a scanner, and are stored in the contents 143 in association with at least the names of the subjects and scenery. In addition, if the electronic device 1 is equipped with an imaging means, the images may be images photographed by the imaging means or images acquired from an external device via a communication means. The electronic dictionary program 141 is a program executed by the CPU 10, and executes functions such as search as an electronic dictionary. The image analysis program 142 is a program executed by the CPU 10, and performs a predetermined image recognition process on the images included in the contents delivered from the electronic dictionary program 141, and outputs the names of the constituent parts of the object displayed as the image so that they can be displayed in association with the respective constituent parts.

[0016] More specifically, by executing image analysis program 142, CPU 10 calculates the features of an object displayed as an image, compares the calculated features with features registered (learned) in advance, and, based on the comparison results, identifies the category to which the object belongs and the constituent parts that make up the object.

[0017] Furthermore, CPU 10 executes image analysis program 142 to obtain names of constituent parts of the object displayed as an image from keyword database 144 based on the identified category and constituent parts, and outputs the names of the constituent parts of the object displayed as an image to electronic dictionary program 141 so that they can be displayed in association with each constituent part. CPU 10 executes electronic dictionary program 141 to display the names of the constituent parts delivered from image analysis program 142 in association with the constituent parts of the object displayed as an image.

[0018] The content 143 is made up of a plurality of pre-installed contents (dictionary contents, study contents, or electronic books). In addition, data on a content icon corresponding to the content 143 is also included in the content itself.

[0019] The keyword database 144 stores, for each category of object, the feature amount of the category, the constituent parts of the object of the category, the feature amount of the constituent parts, and the names of the constituent parts in association with each other. For example, if the category is a vehicle, the feature amount of the constituent parts (front bumper, headlights, etc.) in the vehicle and the names of the constituent parts are stored in association with each other.

[0020] The names of the components may be in multiple languages, and may be adapted to the language of the content from which the image is searched. That is, when "car" is searched in an English-related dictionary, if an image is found that is linked to the search result, the image (car) is displayed, and the names of the components of the vehicle shown in the image are displayed in "English." If the dictionary is Japanese-related, the names of the components of the vehicle are displayed in "Japanese." Alternatively, the language may be switched by a specified operation.

[0021] Fig. 2 is a perspective view showing the appearance of the electronic device (electronic dictionary) 1 according to the first embodiment. In Fig. 2, the electronic device (electronic dictionary) 1 is composed of a first housing 20 and a second housing 21, which have a structure that allows the first housing 20 and the second housing 21 to be opened and closed in two by a hinge section 22. The first housing 20 is provided with a display section 12 and a touch panel 13, and the second housing 21 is provided with a key input section 11.

[0022] B. Operation of the First Embodiment Next, the operation of the above-mentioned first embodiment will be described. 3 and 4 are flow charts for explaining the operation of the electronic device (electronic dictionary) 1 of the first embodiment. In the electronic device (electronic dictionary) 1, the CPU 10, with the home screen displayed on the display unit 12 (step S10), judges whether or not a touch operation to any icon has been detected by a user operation (step S12). Then, if it is judged that a touch operation to any icon on the home screen has been detected (YES in step S12), the CPU 10 proceeds to a process corresponding to each icon (step S14).

[0023] Fig. 5 is a schematic diagram showing an example of a display screen (home screen) in the electronic device 1 according to the first embodiment. As shown in Fig. 5, a keyword input field 31 for search, icons for calling various functions, and a list of content icons are displayed on the home screen 30. On the home screen 30, it is possible to perform a cross-search of various contents using a keyword input in the keyword input field 31, obtain search results (meanings, related images, translations, etc.), and display them on a separate screen, or launch content indicated by an icon.

[0024] In the first embodiment, for ease of understanding, the content icons are shown as blank rectangles and distinguished by alphabets, but in reality, a design is made in which the content name, function name, etc. are written against the cover of the content, etc., so that the content or function assigned to the content icon can be visually recognized. In addition, an example of an operation in which, when a keyword is entered in the keyword entry field 31 for search on the home screen 30, the entered keyword is used to search across various contents will be described below, but it is also possible to start a desired content and search within that content.

[0025] On the other hand, if it is determined that a touch operation on an icon on the home screen 30 has not been detected (NO in step S12), the CPU 10 determines whether or not a keyword has been input in the keyword input field 31 of the home screen 30 and a search has been instructed (step S16). If it is determined that a search has been instructed (YES in step S16), the CPU 10 performs a cross-search of various contents by the keyword input in the keyword input field 31 (step S18) and displays the search result on the display unit 12 (step S20).

[0026] 6 is a schematic diagram showing an example of a display screen (search result screen) in the electronic device 1 according to the first embodiment. The example shown in FIG. 6 shows a display screen of the search results when "car" is entered as a keyword in the keyword input field 31 of the home screen 30. In the search result screen 40, a list of words starting with the keyword "car" is displayed in the list area 41, and search results (meanings) corresponding to the highlighted word "car" are displayed in the search result area 42. When a desired word is selected (touched) from the word list in the list area 41, the search results (meanings) corresponding to the selected word are displayed in the search result area 42.

[0027] In this way, the user can view the content (search results; meaning) for the keyword "car". At this time, the content of the search results area 42 can be scrolled up and down by moving the scroll bar on the right side of the search results screen 40 up and down. The user can select a word that the user wants to search further from the content displayed in the search results area 42 as a keyword and perform further dictionary search. In addition, if a "diagram" button 43 indicating that an image is linked to the keyword "car" is displayed in the search results area 42 as shown in FIG. 6, it is possible to display images related to the keyword "car" by selecting the "diagram" button 43.

[0028] With the search result screen 40 displayed, the CPU 10 judges whether or not a search has been instructed by selecting any word in the content as a keyword (step S22). If the CPU 10 judges that the user has selected any word in the content as a keyword and instructed a search (YES in step S22), the CPU 10 returns to step S18, cross-searches the various contents again with the newly selected keyword (step S18), and displays the search result for the new keyword on the display unit 12 (step S20).

[0029] On the other hand, if the CPU 10 determines that the user has not selected a word in the content as a keyword to instruct a search (NO in step S22), it determines whether or not the "Diagram" button 43 in the search results has been selected (step S24). If the CPU 10 determines that the "Diagram" button 43 in the search results has not been selected (NO in step S24), it returns to step S20 and continues to display the search result screen 40 shown in FIG.

[0030] On the other hand, if the CPU 10 determines that the "Graph" button 43 in the search results has been selected (YES in step S24), it displays the corresponding image on the display unit 12 (step S26).

[0031] Fig. 7 is a schematic diagram showing an example of a display screen (image) in the electronic device 1 according to the first embodiment. The example shown in Fig. 7 shows a display screen of an image related to the keyword "car" when the "Diagram" button 43 in the search result is selected from the search result screen 40 shown in Fig. 6. On the image display screen 50, an image (vehicle) linked to the "Diagram" button 43 is displayed in an image display area 51. Also, an "Image analysis" button 52 for instructing analysis of the displayed image is displayed in the image display area 51.

[0032] The user can view images related to the keyword "car" on the image display screen 50. At this time, the user can select (touch) an "Image Analysis" button 52 displayed in the image display area 51. It should be noted that the user can return to the previous screen (i.e., the search result screen 40 shown in FIG. 6) by selecting (touching) a "Back" button 53.

[0033] The CPU 10 judges whether or not the user has selected (touched) the "Back" button 53 on the image display screen 50 (step S28). If the CPU 10 judges that the user has selected (touched) the "Back" button 53 (YES in step S28), the CPU 10 returns to step S20 and continues to display the search result screen 40 shown in FIG.

[0034] On the other hand, if the CPU 10 determines that the user has not selected (touched) the "Back" button 53 (NO in step S28), it determines whether or not the user has selected (touched) the "Image Analysis" button 52 (step S30). If the CPU 10 determines that the user has not selected (touched) the "Image Analysis" button 52 (NO in step S30), the process returns to step S26 and the image display screen 50 shown in FIG. 7 continues to be displayed.

[0035] On the other hand, when the CPU 10 determines that the user has selected (touched) the "Image Analysis" button 52 (YES in step S30), it passes the image displayed in the image display area 51 to the image analysis program 142 (step S32). The CPU 10 executes the image analysis program 142, analyzes the image displayed in the image display area 51, and calculates the feature amount of the object (step S34). Next, the CPU 10 compares the calculated feature amount with the feature amount registered (learned) in advance (step S36), and identifies the category to which the object belongs and the constituent parts that make up the object based on the comparison result (step S38).

[0036] Specifically, image processing such as edge enhancement and contrast enhancement is performed on the image, and the category to which the object belongs is identified based on the feature values ​​(shape, color, contrast, etc.) of the object in the image. In other words, it is identified to which category the object belongs, such as vehicles or ships. Then, it is identified which part of the image corresponds to which component part of the object in the identified category, based on the feature values ​​(shape, relative positional relationship, color, contrast, etc.) of the components that make up the object.

[0037] Next, the CPU 10 acquires the names of the constituent parts of the object contained in the image from the keyword database 144 based on the identified category and constituent parts (step S40), and outputs the acquired names (constituent part name list 60) so that they can be associated with the constituent parts in the image (step S42).

[0038] Fig. 8 is a conceptual diagram showing an example of a component part name list 60 output from the image analysis program 142 by the electronic device 1 of the first embodiment. As shown in Fig. 8, for example, when the object of the image is identified as a vehicle (automobile), names of the components constituting the vehicle, such as the front bumper, headlights, and turn signals, are obtained, and a component part name list 60 in which the component parts in the image and their names are associated with each other is output. After that, the image analysis program 142 is terminated.

[0039] CPU 10 determines whether or not the names of the constituent parts have been received from image analysis program 142 (step S44), and waits until the names of the constituent parts have been received from image analysis program 142 (NO in step S44). On the other hand, when CPU 10 receives the names of the constituent parts (constituent part name list 60) from image analysis program 142 (YES in step S44 in FIG. 4), it displays the acquired names of the constituent parts in association with the objects of the image on image display screen 50 (step S46).

[0040] 9 is a schematic diagram showing an example of a display screen (image) of the image recognition result by the electronic device 1 of the first embodiment. When the CPU 10 receives the names of the constituent parts (constituent part name list 60) from the image analysis program 142, it displays the names of the constituent parts in association with each part of the object displayed as an image in the image display area 51 on the image display screen 50, as shown in FIG.

[0041] In this way, the user can view not only the images linked to the keywords, but also the names of the constituent parts that make up the object displayed as an image. According to the first embodiment, the names of the constituent parts do not need to be stored in advance as data in the content for each image. For example, even if the content includes images of various vehicles, the names of the constituent parts can be displayed in association with the constituent parts of the vehicles in the images without having to register the names of the constituent parts in advance for each image (vehicle).

[0042] The CPU 10 judges whether or not the user has selected (touched) the "Back" button 53 on the image display screen 50 on which the names of the component parts are displayed (step S48). If the CPU 10 judges that the user has selected (touched) the "Back" button 53 (YES in step S48), the CPU 10 returns to step S20 and continues to display the search result screen 40 shown in FIG.

[0043] On the other hand, if the CPU 10 determines that the user has not selected (touched) the "Back" button 53 (NO in step S48), it determines whether or not the user has selected (touched) the name of a component part (step S50). If the CPU 10 determines that the user has not selected (touched) the name of a component part (NO in step S50), it returns to step S48 and continues to display the image display screen 50 shown in FIG.

[0044] On the other hand, if CPU 10 determines that the user has selected (touched) the name of a component part (YES in step S50), it returns to step S18, cross-searches various contents again using the selected name of the component part as a keyword (step S18), and displays the search results for the new keyword on display unit 12 (step S20).In this way, by selecting (touching) the name of a component part of an object included in an image, the user can further search using the name of the component part as a keyword, and even for an object such as an image for which it is difficult to input keywords for a search, a dictionary search based on the object can be performed.

[0045] Fig. 10 is a schematic diagram showing an example of a display screen (image) of a search result further searched from the display screen (image) of the image recognition result by the electronic device (electronic dictionary) 1 of the first embodiment. For example, when the user selects (touches) the name of the component part "windshield" on the image display screen 50 shown in Fig. 9, a search result 70 searched using "front glass" as a keyword is displayed on the display unit 12 as shown in Fig. 10. Even when "headlight", "door", "tailgate", etc. are selected (touched) as keywords, search results 71 to 73 similarly searched are displayed on the display unit 12.

[0046] In addition, if a search is performed using the names of constituent parts of an object contained in an image as keywords and the search results include a new image, the above-mentioned image analysis can also be performed on this new image to display the names of the constituent parts contained in the new image, and the names of the constituent parts can then be searched for as keywords.

[0047] According to the first embodiment described above, in the electronic device (electronic dictionary) 1, when an image is included in the content, the names of the constituent parts of the object included in the image are obtained and displayed in association with the constituent parts of the object in the image, and further searches are made possible using the names of the constituent parts as keywords. Therefore, even for objects such as images, for which it is difficult to set keywords for searches, keywords can be generated based on the object included in the image.

[0048] C. Second embodiment Next, a second embodiment of the present invention will be described. Fig. 11 is a conceptual diagram showing a system configuration of a second embodiment of the present invention. In Fig. 11, the second embodiment assumes a case where a WEB dictionary provided by a WEB service providing server 84 on the Internet 83 is used from an electronic device such as a smartphone 80 or a personal computer 81. Hereinafter, electronic devices such as the smartphone 80 and the personal computer 81 are collectively referred to as electronic device 82. The electronic device 82 uses an existing WEB browser to access the WEB service providing server 84 via the Internet 83 and uses the WEB dictionary provided by the WEB service providing server 84.

[0049] The WEB service providing server 84 has a function equivalent to the image analysis program 142 of the first embodiment described above, in addition to a function of providing a search service using the above-mentioned WEB dictionary to the electronic device 82. The WEB service providing server 84 also includes a keyword database 85. The keyword database 85 is the same as the keyword database 144 of the first embodiment described above, and therefore a description thereof will be omitted.

[0050] 12 is a schematic diagram showing an example of a display screen (search result screen) of the electronic device (smartphone, personal computer) 100 according to the second embodiment. The electronic device 82 uses an existing WEB browser to access a WEB service providing server 84 via the Internet 83, and inputs a keyword "car" into a keyword input field 91 for searching a specific WEB page 90 provided by the WEB service providing server 84 to instruct a search.

[0051] When a user instructs a search on WEB page 90, WEB service providing server 84 displays a list of words beginning with the keyword "car" entered in keyword input field 91 in list area 92, and also searches a specified database (not shown) for the highlighted word "car" and updates WEB page 90 to display the search results (meaning) in search result area 93.

[0052] As in the first embodiment, the contents of the search result area 93 can be scrolled up and down by moving the scroll bar on the right side of the WEB page 90 up and down. The user can select and search for a word to be further searched for from the contents displayed in the search result area 93. In addition, if an image is associated with the keyword "car" in the contents of the contents displayed in the search result area 93, the image can be displayed by selecting the "Diagram" button 94.

[0053] Fig. 13 is a schematic diagram showing an example of a display screen (image) in the electronic device 82 according to the second embodiment. Fig. 13 shows an example of a display of a WEB page 100 when a "Diagram" button 94 in content is selected. An image (vehicle) linked to the "Diagram" button 94 is displayed on the WEB page 100. In addition, an "Image Analysis" button 101 for instructing analysis of the displayed image is displayed on the WEB page 100.

[0054] FIG. 14 is a schematic diagram showing an example of a display screen (image) of the image recognition result by the electronic device 82 of the second embodiment. When the "image analysis" button 101 on the WEB page 100 is selected (touched), the WEB service providing server 84 analyzes the image and calculates the feature amount of the object. Next, the WEB service providing server 84 identifies the category to which the object belongs and the constituent parts that constitute the object based on the result of comparing the calculated feature amount with the feature amount registered (learned) in advance. Furthermore, the WEB service providing server 84 acquires the names of the constituent parts that constitute the object contained in the image from the keyword database 85 based on the identified category and constituent parts, and updates the WEB page 110 so as to display the acquired names of the constituent parts in association with the object of the image, as shown in FIG. 14.

[0055] 14, the names of the constituent parts are displayed in association with each part of the object displayed as an image, and a list of the names of the constituent parts is displayed in a name list area 111 on the right side. In the example shown, the name of the constituent part "emblem" in the name list area 111 is shown as being selected (touched).

[0056] 15 is a schematic diagram showing an example of a display screen (image) of a search result further searched from the display screen (image) of the image recognition result by the electronic device 82 of the second embodiment. When the user selects (touches) the name of a component part in the name list area 111, the WEB service providing server 84 searches again using the name of the selected component part as a keyword, and updates the WEB page 120 to display the search results for the new keyword.

[0057] According to the second embodiment described above, a user can select (touch) the name of a component part of an object contained in an image and further search using the name of the component part as a keyword. Even for objects such as images, for which it is difficult to set keywords for searching, keywords based on the object contained in the image can be generated.

[0058] In the second embodiment, the image analysis is performed by the WEB service providing server 84, but this is not limited to the above, and the names of the component parts may be obtained using another server (image analysis server) on the Internet.

[0059] In addition, in the first and second embodiments described above, the names of the constituent parts that make up the object of the image have been used as examples of keywords related to an image, but this is not limited to this, and any wording related to the object of the image may be used, for example, a list of parts (names, model numbers, etc.) used in the object, or a list of part manufacturers (addresses, URLs, telephone numbers, distributors, etc.).

[0060] In addition, in the first embodiment described above, the image analysis program 142 in the electronic device (electronic dictionary) 1 performs image analysis, but this is not limited to the above, and the electronic device may be provided with a communication function and may acquire names of component parts (keywords related to the object in the image) using a server (image analysis server) on the Internet.

[0061] Furthermore, in the first and second embodiments described above, the keyword databases 85 and 144 may be machine-learned databases using image recognition technology by artificial intelligence (AI), and names indicating constituent parts of an object contained in an image (keywords related to the object in the image) may be acquired based on the result of comparing the feature amount calculated by image analysis with the machine-learned database. Furthermore, data associating the feature amount calculated by image analysis with the acquired names indicating the constituent parts (keywords related to the object in the image) may be used as training data to train the database by machine learning. [Explanation of symbols]

[0062] 1...electronic device (electronic dictionary), 10...CPU, 11...key input section, 12...display section, 13...touch panel, 14...memory, 141...electronic dictionary program, 142...image analysis program, 143...content, 144...keyword database, 20...first housing, 21...second housing, 22...hinge section, 30...home screen, 31, 91...keyword input field, 40...search result screen, 41, 92...list area, 42, 93...search result area, 43, 94..."Diagram" button, 50...Image display screen, 51...Image display area, 52, 101..."Image analysis" button, 53..."Back" button, 60...Component part name list, 70-73...Search results, 80...Smartphone, 81...Personal computer, 82...Electronic device, 83...Internet, 84...WEB service providing server, 85...Keyword database, 90, 100, 110...WEB page, 111...Name list area

Claims

1. an input unit for inputting an image; Acquire features of the input image; generating keywords to be used as search targets that are related to the object included in the input image based on the acquired feature amount; A control unit; An electronic device comprising:

2. The control unit is comparing the acquired feature amount with feature amounts of at least one object registered in advance, and identifying the object included in the input image based on the comparison result; generating keywords related to the identified objects; 2. The electronic device according to claim 1 .

3. The control unit is If a new image is included in the result of searching the predetermined content using the generated keyword, the new image is input to the input unit and searched again.

2. The electronic device according to claim 1 .

4. The method further includes a machine learning database for generating keywords related to the object included in the input image based on the calculated feature amount, The control unit is generating keywords related to the object included in the input image based on a result of comparing the acquired feature amount with the database; 2. The electronic device according to claim 1 .

5. The control unit is and performing machine learning on the database using data in which the acquired feature quantities and the generated keywords are associated with each other as training data.

5. The electronic device according to claim 4.

6. The control unit is 2. The electronic device according to claim 1, wherein, when a plurality of keywords are generated, the plurality of keywords are displayed in a list on a display unit.

7. A search method executed by a search device, comprising: A step of acquiring a feature amount of an input image; generating keywords to be searched for that are related to the object included in the input image based on the acquired feature amount; A keyword generating method comprising:

8. On the computer, A feature acquisition function that acquires the features of the input image; a keyword generation function for generating keywords to be used as search targets related to objects included in the input image based on the acquired feature amount; A program to achieve this.

Images