CN103377270A - Image searching method and system - Google Patents

Abstract

An image search method and system can find related target images from multiple images through an initial image. A user selects an initial object in the initial image, and then the image search system obtains the initial keyword corresponding to the initial object, and then obtains the related target image based on the initial keyword.

Description

Image search method and system

technical field

The present invention relates to an image search method and system; in particular, to an image search method and system capable of finding at least one associated target picture from multiple pictures through an initial picture.

Background technique

In the existing network search method, after the user inputs a keyword, the search engine returns data according to the keyword. This method is mostly based on text search. The search engine compares various data according to keywords, identifies the corresponding web pages, documents or images and provides them to users.

Text-based searches work well for text-based documents such as articles, news, or books, but not for image documents. In the known image search methods, for example, US Patent No. 6,859,802 uses user feedback to perform image search. The specific method is that if the user wants to search for images related to "grape", the user first uploads several pictures with "grape" to the system, and the system analyzes the common characteristics of these pictures (such as boundary value, color or texture, etc.) ), it is found that they all have the feature of "grape", so the images containing "grape" can be retrieved and provided to the user. This method relies heavily on the ability to distinguish image features, and it is easy to fail to analyze the common features of images.

In addition, the image search system disclosed in Taiwan Patent Publication No. 201117027 searches for a more accurate comparison image based on the overlap rate, contour matching rate, and color matching rate of the original image and the comparison image. However, it is well known that the features of an image are not limited to the overlapping rate, contour matching rate and color matching rate; simply searching with these features may not necessarily find the image that the user wants.

Therefore, it is necessary to provide an image search method and system to solve the deficiencies in the prior art.

Contents of the invention

The main purpose of the present invention is to provide an image search method, which has the effect of finding at least one associated target image from multiple images through an initial image.

Another main objective of the present invention is to provide an image search system, which has the effect of finding at least one associated target image from a plurality of images through an initial image.

In order to achieve the above object, according to an embodiment of the present invention, the image search method of the present invention uses an initial picture to search for at least one associated target picture from multiple pictures, wherein each of the multiple pictures contains at least A clickable object, and each clickable object corresponds to at least one keyword. The image search method includes: receiving a selection of an initial object in the initial image, the initial object being a selectable object in at least one selectable object included in the initial image; obtaining the initial object from the database according to the initial object Corresponding to at least one initial keyword, the at least one initial keyword is the at least one keyword corresponding to the initial object; according to the at least one initial keyword, the at least one target picture is obtained from the database; and the at least one target picture is provided .

According to an embodiment of the present invention, the comparison module is obtained by comparing the at least one initial keyword with the at least one keyword corresponding to the at least one clickable object contained in each of the plurality of pictures to obtain The at least one target picture.

According to an embodiment of the present invention, the ranking of the at least one target image is from those containing more matching objects to those containing less matching objects, wherein the matching objects refer to at least one of the corresponding at least one keyword Objects matching the at least one initial key.

According to an embodiment of the present invention, the image search method further includes: receiving a search keyword; and providing at least one corresponding picture according to the search keyword, and the initial picture is one of the at least one corresponding picture.

According to an embodiment of the present invention, the image search system includes a database, a processing module, an indexing module, a comparing module and a providing module. The database is used to store a plurality of pictures, wherein each picture contains at least one clickable object, and each clickable object corresponds to at least one keyword. The processing module is used for receiving a selection of an initial object in an initial image among the plurality of images, wherein the initial object is a selectable object in the at least one clickable object included in the initial image. The indexing module is used for obtaining at least one initial keyword corresponding to the initial object from the database according to the initial object, wherein the at least one initial keyword is the at least one keyword corresponding to the initial object. The comparison module is used for obtaining at least one target picture from the database according to the at least one initial keyword. The providing module is used for providing the at least one target picture.

According to an embodiment of the present invention, the comparison module is obtained by comparing the at least one initial keyword with the at least one keyword corresponding to the at least one clickable object contained in each of the plurality of pictures to obtain The at least one target picture.

According to an embodiment of the present invention, the module is provided for further sorting the at least one target picture, wherein the sorting of the at least one target picture is arranged from those containing more matching objects to those containing less matching objects, wherein the matching objects Refers to at least one of the corresponding at least one keyword that matches the at least one initial keyword.

According to an embodiment of the present invention, the processing module is further configured to receive a search keyword, and provide at least one corresponding picture according to the search keyword, and the initial picture is one of the at least one corresponding picture

The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments, but not as a limitation of the present invention.

Description of drawings

FIG. 1 is a schematic diagram of the usage environment of an image search method according to an embodiment of the present invention;

FIG. 2 is a system architecture diagram of an image search system according to an embodiment of the present invention;

Fig. 3 is a schematic diagram of an image according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a database according to an embodiment of the present invention;

FIG. 5 is a flowchart of steps of an image search method according to an embodiment of the present invention;

FIG. 6 is a schematic diagram of an image search page according to an embodiment of the present invention;

FIG. 7 is a schematic diagram of a corresponding picture according to an embodiment of the present invention;

FIG. 8 is a schematic diagram of a target image according to an embodiment of the present invention.

Among them, reference signs

Image Search System 1 Database 10

image number 11 object number 12

Object number 13 Keyword 14

Processing Module 20 Indexing Module 30

Compare module 40 Provide module 50

Image 101 Objects 101a, 101b, 101c, 101d

Image 102,103,104,105,106 User 110

Computer 112 Screen 114

Network 120 Image Search Page 600

Keyword input area 601 Search key 602

Detailed ways

In order to make the above and other objects, features and advantages of the present invention more comprehensible, specific embodiments of the present invention are listed below and described in detail in conjunction with the accompanying drawings.

Please refer to FIG. 1 below, which is a schematic diagram of an operating environment of an image search method according to an embodiment of the present invention. As shown in FIG. 1 , a user 110 uses a computer 112 to connect to the image search system 1 through a network 120 to perform image search, and the image search system 1 provides the searched target images and sends them back to the computer 112 .

According to an embodiment of the present invention, the computer 112 is a general desktop computer, but the present invention is not limited thereto; it can also be any electronic device that can be connected to the network such as a notebook computer, a smart phone, or a tablet computer. It should also be noted that the computer 112 may also include other more detailed modules or components.

Please refer to FIG. 2 , which is a system architecture diagram of an image search system according to an embodiment of the present invention. As shown in FIG. 2 , according to an embodiment of the present invention, the image search system 1 of the present invention includes a database 10 , a processing module 20 , an indexing module 30 , a comparing module 40 and a providing module 50 . It should be noted that, in addition to being configured as a hardware device, software program, firmware or a combination thereof, each of the above modules can also be configured by means of a circuit loop or other appropriate types; and, in addition to being configured individually, each module can also be combined type configuration. In addition, this embodiment is only an example of a preferred embodiment of the present invention, and all possible combinations of changes are not described in detail in order to avoid redundant description. However, those skilled in the art should understand that not all the above-mentioned modules or components are necessary. And in order to implement the present invention, it may also include other more detailed existing modules or components. Each module or element may be omitted or modified as required, and there may not be other modules or elements between any two modules.

Next, please refer to FIG. 3 and FIG. 4 , which are respectively a schematic diagram of an image and a schematic diagram of a database according to an embodiment of the present invention. According to an embodiment of the present invention, multiple images are stored in the database 10, and each image is analyzed into multiple objects, and each object corresponds to at least one keyword. As shown in FIG. 3 , according to an embodiment of the present invention, after analysis, the image 101 includes four objects 101 a , 101 b , 101 c and 101 d , which are respectively marked with oval frames. How to analyze the objects included in the image is within the scope of the prior art. For example, it can be automatically judged by the program according to the boundary value, color or texture, or the objects can be marked manually, and will not be repeated here.

As shown in FIG. 4 , the database 10 is divided into a plurality of columns, respectively recording relevant information of each image. According to an embodiment of the present invention, the image number 11 field records the index of each image. For example, image 101 is the first image in the database 10, so its number is 101; the numbers of other images can be deduced accordingly.

The number of objects 12 field records the number of objects contained in each image. If the image 101 contains 4 objects after analysis, the number of objects is 4.

The object number 13 field records the object number of each object of each image. For example, the image 101 contains 4 objects after analysis, and their numbers are respectively 101a, 101b, 101c and 101d.

The keyword 14 field records the corresponding keywords of each object of each image. For example, the corresponding keywords of the object 101a are "helmet" and "head". It should be noted that there can be one or more keywords corresponding to the object. Regarding how to set the corresponding keyword of the image object, it can be manually set, but the present invention is not limited thereto.

Next, please refer to FIG. 5 , which is a flowchart of steps of an image search method according to an embodiment of the present invention; and please refer to FIG. 2 and FIGS. 6 to 8 , which relate to a specific embodiment of the present invention. The image search method of the present invention will be described below with the image search system 1 shown in FIG. 2 . However, it should be noted that the image search method of the present invention is not limited to the image search system 1 shown in FIG. 2 .

First go to step 501: receiving a search keyword.

Please refer to FIG. 6 , which is a schematic diagram of an image search page according to an embodiment of the present invention. According to an embodiment of the present invention, the user uses the computer 112 to connect to the image search system 1 to search for images. The computer 112 includes a screen 114 for displaying images.

As shown in FIG. 6 , when the computer 112 is connected to the image search system 1 , an image search page 600 is displayed on the screen 114 , which includes a search keyword input area 601 and a search key 602 . The user can input the search keyword in the keyword input area 601 and press the search key 602 to send the search keyword to the image search system 1 , and the image search system 1 receives the search keyword through the processing module 20 . For example, in this example, the user entered "playing baseball" as the search keyword.

Then proceed to step 502: providing corresponding pictures according to the search keywords.

Please refer to FIG. 7 , which is a schematic diagram of corresponding pictures according to an embodiment of the present invention. After the image search system 1 receives the search keyword, the processing module 20 can further provide at least one corresponding picture according to the search keyword.

As shown in Figure 7, the processing module 20 searches the database 10 for images 101, 102, 103 and 104 related to "playing baseball" according to the search keyword "playing baseball" as corresponding pictures, and sends them to the computer 112 through the network 120 While displayed on the screen 114, the user can select one of them as the initial picture. It should also be noted that how to provide corresponding images according to the search keywords is within the scope of the prior art and should be well known to those skilled in the present invention, and will not be repeated here.

Then proceed to step 503: receiving a selection of an initial object in the initial picture.

Please refer to FIG. 7, the user can select one of the images 101, 102, 103 and 104 as the initial image and click one of the objects as the initial object, and send it to the image search system 1, and the image search system 1 passes the processing module 20 to receive the selection.

According to an embodiment of the present invention, when the processing module 20 provides the corresponding pictures according to the search keywords, the objects included in each corresponding picture are displayed on the screen 114 in the form of clickable objects at the same time, and are displayed in an oval frame Indicates clickable objects. Therefore, the user can directly click on an object, and the processing module 20 will identify the object as the initial object after receiving the click instruction, and the image containing the object will be identified as the initial image. For example, in this embodiment, the user clicks on the object 101d of the image 101, and the processing module 20 recognizes the object 101d as the initial object and the image 101 as the initial picture after receiving the click instruction. It should be noted that how to present objects in the form of clickable objects is in the scope of the prior art and should be well known to those skilled in the present invention, and will not be repeated here.

Then proceed to step 504: obtain at least one initial keyword corresponding to the initial object from the database according to the initial object.

After the image search system 1 receives the instruction to select the initial object, it can use the index module 30 to obtain the initial keyword corresponding to the initial object from the database 10 according to the initial object, where the initial keyword is the corresponding initial object. keywords.

Please refer to FIG. 4 , after determining that the object 101d is the initial object, the indexing module 30 may obtain the keyword "bat" corresponding to the object 101d from the database 10 as the initial keyword. It should be noted that the initial keywords are not limited to only one set. If there are multiple sets of keywords corresponding to the initial object, all corresponding keywords will be used as the initial keywords.

Then proceed to step 505: obtain at least one target picture from the database according to the at least one initial keyword.

After the image search system 1 obtains the initial keyword, the comparison module 40 can then obtain the target picture from the database 10 according to the initial keyword. According to an embodiment of the present invention, the comparison module 40 obtains the target image by comparing the initial keyword with the keywords corresponding to the clickable objects contained in the plurality of images stored in the database 10 .

Please refer to FIG. 8 , which is a schematic diagram of a target picture according to an embodiment of the present invention. In this embodiment, the images 105, 106 and 102 are found by comparing the initial keyword "bat" with the keywords corresponding to the clickable objects contained in the multiple pictures stored in the database 10. The keywords corresponding to the clickable objects included in the three images also include "ball bat", so the three images 105, 106 and 102 are used as target images.

Finally, go to step 506: provide the at least one target picture.

Finally, the providing module 50 provides the target picture, for example, transmits the target picture to the computer 112 via the network 120 and displays it on the screen 114 . According to an embodiment of the present invention, the providing module 50 can further sort the target pictures, from those containing more matching objects to those containing less matching objects, wherein a matching object refers to at least one of the corresponding keywords and Objects matching the initial key.

Please refer to FIG. 8, images 105, 106 and 102 contain 3, 2 and 1 matching objects respectively, and the keywords corresponding to these matching objects also have "ball bat", so the three images of images 105, 106 and 102 are used as target image. According to an embodiment of the present invention, the target image is based on the number of matching objects contained therein. Therefore, as shown in FIG. 8 , the image 105 containing 3 matching objects is displayed at the top, the image 106 containing 2 matching objects is displayed next, and the image 102 containing 1 matching object is displayed last. However, it should be noted that the method for sorting the target pictures in the present invention is not limited thereto.

According to an embodiment of the present invention, as shown in FIG. 8 , when the target picture is displayed, the matching objects included in the target picture will be marked with an oval frame to indicate the position of the matching objects. However, the image search method of the present invention is not limited thereto, and the position of the matching object may also be marked in other ways, or may not be marked.

It should be noted here that the image searching method of the present invention is not limited to the order of the above steps, as long as the purpose of the present invention can be achieved, the order of the above steps can also be changed.

Of course, the present invention can also have other various embodiments, and those skilled in the art can make various corresponding changes and deformations according to the present invention without departing from the spirit and essence of the present invention, but these corresponding Changes and deformations should belong to the scope of protection of the appended claims of the present invention.

Claims (8)

1. image method for searching, in a plurality of pictures, seek at least one Target Photo that is associated by an initial picture, wherein each picture in these a plurality of pictures respectively comprises at least one object that clicks, and each can click object respectively corresponding at least one key word, it is characterized in that this image method for searching comprises:

Reception is to the selection of the initial object of one in this initial picture, this initial object comprise for this initial picture this at least onely click one in the object and can click object;

By an index module, in a database, obtain the corresponding at least one initial key word of this initial object according to this initial object, this at least one initial key word is corresponding this at least one key word of this initial object;

By a comparing module, in this database, obtain this at least one Target Photo according to this at least one initial key word; And

This at least one Target Photo is provided.

2. image method for searching according to claim 1, it is characterized in that, this comparing module be by compare that this at least one initial key word and this a plurality of picture respectively comprised this at least one click object each this corresponding at least one key word, to obtain this at least one Target Photo.

3. image method for searching according to claim 2, it is characterized in that, the ordering of this at least one Target Photo, be to arrange to comprising the less object that conforms to from comprising the more object that conforms to, wherein this object that conforms to refers at least one object that conforms to this at least one initial key word in corresponding this at least one key word.

4. image method for searching according to claim 1 is characterized in that, also comprises:

Receive one and search key word; And

Search key word according to this, at least one corresponding picture is provided, this initial picture is one of them of this at least one corresponding picture.

5. an image search system is characterized in that, comprising:

One database, in order to store a plurality of pictures, wherein each picture respectively comprises at least one object that clicks, and each can click object respectively corresponding at least one key word;

One processing module, in order to receive the selection to the initial object in the initial picture in these a plurality of pictures, wherein this initial object comprise for this initial picture this at least onely click one in the object and can click object;

One index module, in order to obtain the corresponding at least one initial key word of this initial object according to this initial object in this database, wherein this at least one initial key word is corresponding this at least one key word of this initial object;

One comparing module is in order to obtain at least one Target Photo according to this at least one initial key word in this database; And

One provides module, in order to this at least one Target Photo to be provided.

6. image search system according to claim 5, it is characterized in that, this comparing module be by compare that this at least one initial key word and this a plurality of picture respectively comprised this at least one click object each this corresponding at least one key word, to obtain this at least one Target Photo.

7. image search system according to claim 6, it is characterized in that, this provides module further in order to this at least one Target Photo of ordering, the wherein ordering of this at least one Target Photo, be to arrange to comprising the less object that conforms to from comprising the more object that conforms to, wherein this object that conforms to refers at least one object that conforms to this at least one initial key word in corresponding this at least one key word.

8. image search system according to claim 5 is characterized in that, this processing module is further searched key word in order to receive one, and provides at least one corresponding picture according to this search key word, and this initial picture is one of them of this at least one corresponding picture.

Images