CN107153666B

CN107153666B - Picture searching method

Info

Publication number: CN107153666B
Application number: CN201610137859.6A
Authority: CN
Inventors: 魏立江
Original assignee: Individual
Current assignee: Individual
Priority date: 2016-03-02
Filing date: 2016-03-02
Publication date: 2022-03-04
Anticipated expiration: 2036-03-02
Also published as: CN107153666A

Abstract

A searching method using picture as searching code, which is called picture searching method, solves the defects of two-dimension code, bar code and link code, and makes the target content of picture be easily identified by public, and can automatically read the information of picture by image input device or photoelectric scanning device. After a picture to be processed is recorded into a system, a background with default shape and color by the system is added to the picture, a main pattern in a background outline range is coded in a minimum vector diagram format and is set to point and then is filed, and simultaneously, the picture containing the background before coding is output and edited in the main stream format for a user to use.

Description

Picture searching method

The invention relates to a searching method using pictures as searching codes, which is called as a picture searching method.

Technical background the current two-dimensional codes, bar codes, links and other search codes have the defects of being not visual and beautiful enough, the designated target content is not easy to be directly identified by the public, and although some marks or patterns are added into some two-dimensional codes, certain limitation still exists in the aspect of information transmission.

The invention aims to provide a generating and searching method, which takes a plane picture as a carrier, encodes the picture according to a specific range and format and establishes a targeted connection with the content indicated by the picture, so that people can visually identify the meaning represented by the picture through the picture and can automatically identify the information indicated by the picture through an input method and a search engine or through an image input device or a photoelectric scanning device or a digital terminal device. Here, we refer to the code in which pictures are coded in a specific format as picture code.

In the drawing explanatory characters, (fig. 1) is a general working block diagram of the invention, wherein a broken line square frame is a picture code generation process block diagram, and a broken line square frame is a picture scanning and retrieving working block diagram. (FIG. 2) is a symbolic legend representing the word "shopping" in TTF format (format of fonts in input method) designed in the manner of simple strokes. (FIG. 3) the main pattern is a symbolic legend of the word "love team" in TTF format, and the background pattern in the pattern is a system background given by the system, and the background is output together with the main pattern after editing is completed. (FIG. 4) is a legend for the word "shopping cart" in AI format (one of the formats commonly used for vector graphics), with the dashed boxes indicating the background outline locations of the figure. (FIG. 5) the dotted line square frame is the three layer pattern of the symbol legend of "shopping" under the GIF format (dynamic graph format), and the continuous output of the three layers can obtain the dynamic effect graph of rolling up the dynamic trolley wheels.

The invention will be further explained with reference to the drawings.

The object of the invention is achieved by: in (fig. 1), the part inside the dashed line box is the production generation process of the picture code. Firstly, a pattern is selected or designed for contents to be specified, the contents to be specified can be characters, can be pictures, audios and videos, can also be various contents such as links and the like, the selected or designed pattern can preferably represent the contents, then the pictures are loaded into a system to be preprocessed, an editing link is firstly entered, the system can give out a pure white rectangular block with clear outline as a background when the pictures are edited, after the pictures are loaded into the system, a user can edit the pictures according to needs and system prompts, the pictures can be cut, characters or characters and the like can be added. The picture represents a certain meaning, in order to facilitate retrieval, a name can be defined for the picture, a unique corresponding keyword is set, the unique corresponding keyword is called as a first keyword, the setting of the first keyword follows the principle of uniqueness, the first keyword only corresponds to one unique database picture, the first keyword has the principle of uniqueness and exclusivity, in order to facilitate operation, other words with association relation can be designated for the picture as secondary keywords which are classified as second, third, fourth. After the first keyword is set, the system database is identified, and the first keyword can be effective only after being confirmed to be unique. The first keyword is identified as a prior registration principle. Meanwhile, in order to facilitate retrieval and management of the pictures, the system also determines a single unique code consisting of numbers or characters for the pictures entering the system database, so that the picture content can be directly obtained by inputting the first keyword or the unique code of the pictures after entering the picture searching function, and the unique code is directly granted by the system or can be obtained by user definition and then is confirmed by the system. The first key words and the unique codes of the pictures are selected and output by a user in the picture making process, and the default position of the system is the bottom edge outside the background. The layers of the dynamic graph share a first keyword and a unique code. After the picture is edited, the system generates a synthetic picture with a white rectangle provided by default of the system as a background, and the distance between each side of the rectangle and one point of the outermost edge of each of the upper, lower, left and right sides of the main pattern adopts a default value, as shown in (fig. 4), the area of the dotted line box represents a white rectangular background area, and the distance between the four contour lines and the main pattern adopts a default value, so that a user can customize the synthetic picture according to needs. The key words and the unique coding function are selectively used, and the user can select the setting according to the requirement.

After the editing process is finished, the system starts to encode and generate the picture, the step firstly needs to determine the value range and direction of the picture, and the system can determine the acquisition direction and range of the pattern by adopting a background internal value taking method according to picture information, wherein the background internal value taking method is a method for acquiring effective component information of the picture which is surrounded by a background contour line in the front direction and is downwards recorded by locking an azimuth mark by taking the background contour line as a boundary line. As shown in (fig. 4), the system first locks the default dashed box as a range of values, and locks the mark of the "L" shape in the lower left corner circle in the background as the lower left corner mark, which is the default "lower left corner" mark of the system and is a part of the background, and the position is also in the default position, where the circle does not exist in the actual application, and is used only for auxiliary annotation. The determination of the image acquisition direction is realized by adding an orientation mark or a default orientation background, for example, in the background (fig. 3), because four edges of the background outline are different, the system takes the edge state of the current outline as the default current direction of the value to record all effective components forming the 'love team' logo pattern in the background range. When the system reads the pattern, the background information is stored in the background library of the system, so that the background can be automatically removed, and only the information of the pattern in the background is recorded. The main function of the background is to guide how the system operates, provide orientation and range reference for information acquisition in the picture code generation process, and provide reference for system discovery and picture correction in the retrieval link. The color and shape of the background are given by default of the system, the default background of the system is a white rectangular pattern, the range defined by the outline of the rectangle is used as the default value range of the system, and the proportion of the rectangle is changed along with the proportion change of the main pattern. The background can also be freely set by a user, pictures which accord with uniqueness and background basic conditions after the background color and the background shape are submitted to a system for authentication can be added into a system background library for use by the system, the design of the background needs to follow the principle that the outline is clear, the mark capable of indicating the value range and the direction is provided, and the shape is obvious, the user-defined background adopts the prior registration principle, and is divided into public and exclusive, the same or similar background is removed by examination, the similarity is determined according to a default value, and the background library does not accept the background which has a visual deviation relation with the existing background. As with the custom background, the orientation marker may also be custom-defined by the user, but the shape and placement must be approved by the system for use in the system database. The pattern containing multiple background contours in one picture is based on the outermost complete background contour. And the background added by default or user-defined by the system is output together with the edited main pattern and the orientation mark as well as the first keyword and the unique code in the picture drawing link, and the output format is any or mainstream format.

The coding generation is carried out by adopting a format unified with the picture retrieval, and in order to ensure the uniqueness of the picture code, the system excludes the picture with the shape completely identical with the existing pattern shape of the database, gives the picture a failure, and prompts the resetting. Since the system uses the comparison of whether the shapes and positional relationships of the pattern active ingredients are the same as the number of the active ingredients as the criterion for determining whether the pattern passes, the color of the pattern is not considered. That is, all pictures are converted into black and white default colors after entering the system, and two pictures with the same shape and different colors are regarded as the same picture. For the convenience of storage and retrieval, all recorded pictures adopt a vector diagram format with the minimum volume after the pictures are compressed. The method can carry out edge extraction on a pattern according to a default value through system analysis on the picture, and carry out binarization differentiation to divide the picture into the following two categories: firstly, determining the picture which can be completely binarized within the default value range as a pattern without gradation in a single color, wherein, the main pattern shown in (figure 2) and (figure 3) and the three patterns shown in (figure 5) are all the patterns which are distinct in black and white. For the patterns without gradual change, the TTF format with smaller volume is the default format without gradual change patterns, namely the format generally adopted by characters or symbols in the input method, the format only counts the colored part into the capacity, the background part does not count into the capacity, or the format is converted into the format of the characters or symbols for operation. Secondly, pictures which cannot be completely binarized within the default value range and contain unextractable edge elements are defined as monochrome patterns with gradual change, such as patterns in a dotted line box (figure 4), and the default format of the picture system is an AI format commonly used for vector graphics. The default values mentioned here are the gray value and the saturation value at the time of edge extraction. Since individual patterns may have fade values that happen to be within the default range, the width range values of the edges are also included in the default values. When the picture is read, the picture in the non-vector diagram format is read in the bitmap format and then converted into the vector diagram format for comparison, storage and retrieval, and the picture edge can become fuzzy in the picture reading and transcoding process, so the edge extraction operation is required to be carried out according to a default value.

Because the recording and coding formats of the two types of pictures adopt an algebraic mode to record the set of picture objects, the picture codes after coding are set by an algebraic mode. After the system finishes the vector coding of each object of the picture, a complete algebraic set is uploaded to a system database in a data packet mode for data comparison, the comparison process is carried out according to the recording sequence of the picture objects in a vector diagram format, firstly, the comparison of the first sequence recorded objects is carried out, the data of the first sequence recorded objects are extracted for data comparison of the full database, if the same object data is not found through the comparison, the picture is not repeated in the database, the next link is entered through the comparison, if the same object data is found, the data packet which is completely the same as the first sequence recorded object data is extracted from the database for comparison screening of the second sequence recorded objects, similarly, if the same data is not found, the comparison is entered into the next link, if the same data objects are still found, extracting all data packets containing the object in the sequence to compare the next recording sequence until all the recorded objects are compared, if the repeated information still exists and the data packets recording the repeated information are exactly finished in the same sequence, namely the data of the picture object comprise the same shape, position relation and number, the picture is completely the same as the picture of the database, the picture cannot be added into the database if the repeated information exists, and a prompt that the system authentication cannot pass and the system authentication is required to be reset is given. However, if the same data packet is not found through the last comparison or a certain sequence of comparisons, the comparison process is immediately stopped, which means that no same picture is registered in the database. Since the database registration adopts the prior registration principle, the same picture is rejected.

And immediately entering the next link for the pictures passing the comparison: the pointing direction is set. And meanwhile, the picture can be used by a user, and the picture generated together with the background and the orientation mark, the first keyword and the unique code before coding is used by the user after the output is edited by the system. After entering a pointing link, a single pointing direction can be set for the picture, and information to be represented by the picture is added, wherein the added content comprises documents, pictures, audios and videos, links and the like. Only one target direction can be set for one picture code, but different picture codes can point to the same target. After the setting is finished, the system uploads the generated image code comprehensive data packet to a database of the Internet for storage, and meanwhile, the database also stores the image data and the related data of the background which are generated together with the background after the editing.

For the pictures with dynamic format, as shown in (fig. 5), three pictures in the dashed square frame are three layer patterns of the dynamic picture of the keyword "shopping", and the three layers are continuously output to obtain a picture with dynamic effect. The user can obtain more positive effect by using the dynamic picture to publicize the product. For the pictures in the dynamic format, each layer needs to be treated according to a single picture, each layer needs to be independently coded, compared and stored, the direction setting of each layer also needs to be independently set, and the principle that only one direction can be set for one picture and the directions of different pictures can be the same is still followed. Usually, the directions of layers of a dynamic graph are the same target, so that the target direction of the dynamic graph can be registered as long as the picture of any one of the layers is obtained.

In fig. 1, the part outside the dashed box forms a work flow diagram of the searching end by combining the system database in the dashed box. After a selected picture is loaded into a system in a scanning or digital transmission mode and the like, the system provided with a search end firstly locks a picture outline to find a background, after the guide information of the background is extracted, the pattern is subjected to value coding according to a unified filing coding format and a value range set by the system background when a picture code is made, the visual deviation and distortion of the picture caused by different shooting angles when the picture is loaded into the system are corrected according to default background data or custom background data, the picture is corrected according to the displacement relation and the deformation degree of the background outline stored in the system and the background outline after imaging by comparing and analyzing the background outline in a background database with the loaded picture background outline, and the picture is corrected into a unified front view. The method is characterized in that a step of generating a visual correction picture in a coding mode is completed, and the system records the picture in a vector diagram format when the picture code is manufactured, so that the corrected picture is subjected to coding in the vector diagram format by removing a background in the same format in the next step, and the step is performed by adopting a format and a method which are unified with the picture code manufacturing step. After the coding of the picture is completed, the picture is immediately entered into a picture retrieval link, and the picture which is completely matched with the current picture is found from a system database. The system records the picture in a vector diagram format by adopting an algebraic form to record the picture, so the retrieval method is similar to the comparison link in the process of manufacturing the picture code, except that the same picture is eliminated through comparison in the process of manufacturing the picture code, but the same picture is found and matched at present, and the retrieval elimination is carried out by extracting one by one according to the sequence of the recorded picture objects until the unique corresponding picture is established, and the retrieval is stopped immediately to skip and log in or read the information pointed by the picture. And if the picture code matched with the picture code is not found through searching, giving a prompt without a result. For pictures used in environments prone to contamination, in order to increase the success rate of reading codes, a method of adding a mark allowing reading of approximate values is adopted in editing pictures, as shown in (fig. 3), and the background behind the main pattern is a mark background allowing reading of approximate values. Therefore, once the picture is stained, the system can not find the matched data when searching the picture, and at the moment, because the authorization mark background which can provide approximate values is read when the picture is read, the closest picture can be automatically provided for the user to confirm and then skip to log in. The reading-allowed approximate value mark is a uniform pattern or background which is added on the pattern and is authenticated by the system, the system defaults and customs are both provided, and similarly, the customized reading-allowed approximate value mark is required to pass the system identification for use. In this way, the picture can be read even if it is occluded or partially removed.

Since the first keyword and the unique code are set for the picture in the editing link, after the control with the function is added to the input method, after the picture search button is started, the picture corresponding to the keyword or the unique code after the first keyword or the unique code is input appears at the first option position of the candidate window in the minimum format, and the second, third, fourth.

Under some specific conditions, pictures need to run in other format environments, but all follow the principle that the archiving coding format and the retrieval format are unified, and apply a background representing the specific format for guiding the system to run in which format.

The user can set the information such as product introduction and company website as the search direction of the mark pattern, so that the audience can scan the mark provided by the merchant by using the mobile phone provided with the search end to read or log in the information indicated by the mark pattern, and great convenience is provided for the merchant to carry out propaganda.

The phenomenon of conflict with the system operation is solved by adopting a default value or an authentication channel. The registrant of the picture has the right to delete and modify, subject to the authentication right.

Claims

1. A searching method taking a picture as a searching code is characterized in that a plane picture is taken as a carrier, the picture is acquired and converted into a unified format according to information recorded by the background after a background containing system operation guide information is added to the picture, a keyword of the picture is granted, the uniqueness of the picture is established through comparison with an existing database, a specific target direction and a unique code are given to the picture, the picture is retrieved and matched in the unified format when the picture is filed after the picture is read through image input equipment or photoelectric scanning equipment or digital terminal equipment, and the target direction indicated by the picture is automatically jumped to after the unique corresponding relation is established;

and after the uniqueness of the picture is established by comparing the picture with the existing database, a specific target direction is given to the picture: extracting and comparing the vector diagrams one by one according to the object recording sequence of the vector diagrams, comparing information of the shape, the position relation and the total number of the objects of each vector diagram recorded in a database in a comparison algebraic form to be used as a standard for judging whether the pictures are the same or not, and giving a specific direction to the pictures which are confirmed to be not completely the same as the pattern information of the database;

and (3) converting the picture acquisition into a uniform format according to the information recorded by the background: and determining a range defined by the background outline or a designated area in the background as a value range of the system, determining the current direction of the picture according to the orientation mark, and uniformly converting the filing format of the picture into a minimum black-and-white vector diagram format for operation after removing background information.

2. The method as claimed in claim 1, wherein the picture is added with a background containing system operation guidance information: the image which is clear in outline, provided with a mark capable of indicating a value range and an orientation, obvious in shape and capable of recording the value range, the value direction and the coding format of the image by the system and recording whether the system operation guidance information of the approximate value is allowed to be read is used as the background or part of the background of the main pattern.

3. The method according to claim 1, wherein the range defined by the background contour or the area specified in the background is determined as the value range of the system, and the current direction of the picture is determined according to the orientation flag: the white rectangle is used as the default background of the system, the range defined by the outline of the rectangle is used as the default value range of the system, the proportion of the rectangle is changed along with the proportion change of the main pattern, the custom background and the custom orientation mark are allowed, but the white rectangle can be added into the system database only after being approved by the system.

4. The method as claimed in claim 1, wherein the picture keyword and the unique code are granted: the method comprises the steps of granting a special word corresponding to the picture as a first keyword and granting other related words as secondary keywords, and granting the picture a special unique code.

5. The method as claimed in claim 1, wherein the image is read by an image input device or an optoelectronic scanning device or a digital terminal device and then adopts a format unified with the image archiving: after the picture is read into the system, the picture is compared and analyzed through the background contour in the background database and the loaded picture background contour, the picture is corrected according to the displacement relation and the deformation degree of the background contour stored in the system and the imaged background contour, and the picture is corrected into a uniform front view and then converted into a uniform format during archiving.

6. The method according to claim 1, wherein the planar picture is used as a carrier: the method comprises a single picture and a dynamic picture, wherein all layers of the dynamic picture are treated according to the single picture, and all layers of the dynamic picture share a first keyword and a unique code.

7. A search method using a picture as a search code according to claim 1, characterized in that the picture is read by an image input device or an optoelectronic scanning device or a digital terminal device: the read picture is a picture which is generated by a user without system coding processing after editing and contains a background mark, an orientation mark, a first keyword and a unique code, and the output format is any or mainstream format.

8. The method as claimed in claim 1, wherein the image is read by an image input device, an electro-optical scanning device or a digital terminal device, and then retrieved and matched in a format unified with the image during archiving, and after a unique correspondence is established, the image automatically jumps to the target direction indicated by the image: and after the control is added by the input method, a first keyword or a unique code is input, the option corresponding to the picture can be obtained, and after the option is confirmed, the target direction of the picture is directly jumped and logged in through an upgraded browser application program.