CN106709531B - Method and device for identifying substances used by multi-process matched tobacco materials - Google Patents

Abstract

The invention provides a method and device for identifying substances used in multi-process matching cigarette materials. The method first scans and identifies the information of cigarette materials in paper documents or pictures of tobacco materials; Match with the cigarette material licensing standard database. If the match is successful, the result will be displayed; if the match is unsuccessful, the information of the cigarette material will be matched with the historical experience database. If the match is successful, the result will be displayed. The invention solves the problem that identification and matching are difficult in the prior art due to inconsistent standards and various forms of tobacco materials. The constructed historical experience database with self-learning function, in the case of unsuccessful matching with the cigarette material licensing standard database, is then matched with the historical experience database for a second time, so that the matching accuracy is improved.

Description

A method and device for identifying substances used in multi-process matching cigarette materials

technical field

The invention belongs to the technical field of text recognition and detection, and in particular relates to a method and a device for identifying substances used in multi-process matching cigarette materials.

Background technique

Cigarette materials refer to the materials used in the production of cigarettes, including cigarette paper, stencil printing ink for cigarette paper, filter rod forming paper, cigarette tipping paper, cigarette lining paper, cigarette frame paper, wrapping paper (bars and boxes) ), seal paper, water-based adhesive for cigarette, hot melt adhesive for cigarette, cellulose diacetate tow for cigarette, polypropylene fiber tow for cigarette and triacetin for cigarette.

Substances used in smoking materials refer to the raw materials used in the production process of smoking materials and the substances added to meet the intended use to help improve their quality, characteristics or assist in improving their quality and characteristics, as well as to promote the smooth progress of the production process. It is not a processing aid added to improve the quality and characteristics of the final product.

The use of smoking materials directly affects the safety of cigarette products. Tobacco companies develop lists of permitted substances. All smoking materials entering cigarette products must be reviewed for compliance based on the standards. At present, the audit of licensed substances for cigarette materials adopts the method of manual review, and the compliance review is carried out on the batch substance information of cigarette materials reported by suppliers one by one. Due to the lack of normative material data and other reasons, the review workload is huge, which brings great inconvenience to the reviewers. Moreover, the standards for tobacco materials are not uniform and in various forms, making it difficult to identify and match.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method and device for identifying substances used in multi-process matching tobacco materials, so as to solve the problem that identification and matching are difficult in the prior art due to inconsistent standards and various forms of tobacco materials.

For solving the above-mentioned technical problems, the technical scheme of the present invention is:

The present invention provides an intelligent multi-process matching method for identifying substances used in cigarette materials, including the following method solutions:

Method 1 includes the following steps:

1) Scan and identify information on tobacco materials in paper documents or pictures of tobacco materials;

2) Match the information of the identified smoking materials with the smoking materials licensing standard database, and if the matching is successful, the result will be displayed; wherein, the smoking materials licensing standard database stores the permitted smoking materials using substances. information;

3) If the matching is unsuccessful, the information of the tobacco material is matched with the historical experience database, and if the matching is successful, the result is displayed; wherein, the historical experience database stores the substances used in tobacco materials that have been successfully or unsuccessfully matched before. Information.

The second method, on the basis of the first method, further includes the step of storing the result matched with the historical experience database in the historical experience database.

Method 3 and 4: On the basis of Method 1 and Method 2, respectively, when matching the information of cigarette materials with the historical experience database, formulas are used for the feature set of the information of cigarette materials.

to measure, and display the results according to the matching similarity from high to low; among them, C _i is the center of the feature set cluster, x is the set distribution point of the cluster, k is the number of feature data sets, and E is the optimal value of the result .

Methods 5 and 6 are based on method 1 and method 2, respectively, through OCR identification to identify the tobacco material information in paper documents or pictures of tobacco materials.

The present invention also provides a multi-process matching device for identifying substances used in smoking materials, including the following device solutions:

Device solution one includes the following modules:

A module for scanning and identifying information on tobacco material in paper documents or pictures of tobacco material;

A module used to match the information of the identified smoking materials with the smoking materials licensing standard database, and if the matching is successful, the result will be displayed; wherein, the smoking materials licensing standard database stores the allowed smoking materials using information on substances;

A module used to match the information of tobacco materials with the historical experience database if the matching is unsuccessful, and display the results if the matching is successful; wherein, the historical experience database stores the cigarette materials that have been successfully or unsuccessfully matched before Information on substances used.

The second device solution, on the basis of the first device solution, further includes a module for storing the result matched with the historical experience database in the historical experience database.

Device scheme three and four, based on device scheme one and device scheme two, respectively, when matching the information of cigarette materials with the historical experience database, formulas are used for the feature set of the information of cigarette materials

to measure, and display the results according to the matching similarity from high to low; among them, C _i is the center of the feature set cluster, x is the set distribution point of the cluster, k is the number of feature data sets, and E is the optimal value of the result .

Device solutions 5 and 6 are based on device solution 1 and device solution 2, respectively, through OCR identification to identify tobacco material information in paper documents or pictures.

The beneficial effects of the present invention are:

The method and device for identifying substances used in multi-process matching cigarette materials of the present invention, after identifying the information of cigarette materials, firstly quickly search and query the cigarette material licensing standard library, and match the information of cigarette materials with the standard library, If the matching is unsuccessful, continue to match the information of cigarette materials with the constructed historical experience database to realize intelligent matching and prediction of the text information of cigarette materials used. The method solves the problems of large workload and inaccurate auditing of licensed substances in tobacco materials by manual auditing in the prior art, which can improve the work efficiency of tobacco material auditing personnel and reduce the error rate caused by manual auditing.

Moreover, because the cigarette material licensing standard database may not be able to cover all the chemical formulas, English names, common names, etc. of all cigarette materials, it is often difficult to match the results in one matching process. Therefore, a historical experience database with self-learning function is constructed. , in the case of unsuccessful matching with the cigarette material licensing standard database, a second matching is performed with the historical experience database, so that the matching accuracy is improved.

Description of drawings

Fig. 1 is the main component structure schematic diagram of the present invention;

2 is a schematic diagram of the network connection relationship of the main components of the present invention;

Fig. 3 is the flow chart of the present invention;

Figure 4 is the composition diagram of the paper substance provided by the supplier;

Fig. 5 is the schematic diagram of the licensing standard database of tobacco materials;

Fig. 6 is the schematic diagram of historical experience database;

FIG. 7 is a matching result graph.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

As shown in FIG. 1 , the present invention mainly includes: a handheld device and a server. The handheld device includes a handheld scanning module, a first data transceiver module, and a terminal display module; the server side includes an OCR identification module, a second data transceiver module, an intelligent matching module, a tobacco material licensing standard database module, and a matching experience module.

specific:

The handheld scanning module is used to scan the chemical text information, picture information or text and picture combination information of substances used in cigarette materials, and is also used to confirm the matching result information displayed on the terminal.

The first data transceiver module is used to send the scanned image data sent by the handheld device to the server for intelligent matching processing, and is also used to receive the returned matching result.

The second data transceiver module is used for receiving the picture data information sent from the handheld device, and is also used for sending matching result information to the handheld device.

The OCR identification module is used to scan and identify the picture data sent by the first data transceiver module to identify the specific cigarette material information in the picture.

Tobacco material licensing standard database module, which is used to store the standard information of licensed substances for smoking materials.

The historical experience database module is used to store matching experience data and historical data information.

The terminal display module is used to display the matching result information, and sort and display it according to the similarity of the matching.

FIG. 2 is a schematic diagram of the network connection relationship of the main components of the present invention, mainly including: an intelligent handheld device 1 , a wireless router 2 , an application server 3 and a database server 4 .

specific:

The smart handheld device 1 is used to scan pictures of chemical text information of substances used in cigarette materials, and send the data to the database server 4 through the wireless router 2 through the data transceiver module, and is also used for the operation of confirming the matching results.

The wireless router 2 is used for the smart handheld device 1 to send data to the database server 4, and for the application server 3 to send the comparison matching result information to the smart handheld device.

The application server 3 performs OCR identification on the chemical text information and picture data of the substances used for smoking materials stored in the database server 4, and compares the standard information of the permitted substances for smoking materials stored in the database server 4, and matches the historical experience database information, and passes the comparison result through. The wireless router 2 sends it to the smart handheld device 1.

The database server 4 stores scanned chemical text information pictures of substances used in smoking materials, data sent from the smart handheld device 1 , standard information of licensed substances for smoking materials, matching experience data and historical data information.

As shown in Figure 3, the main flow of the present invention is as follows:

1) Scan the tobacco material paper file line by line through a handheld scanning device, and identify the paper material composition pictures (such as calcium carbonate, CaCO ₃ ) submitted by the material supplier, etc., as shown in FIG. 4 , provided by the supplier in this embodiment composition chart of paper substances.

2) The scanning device sends the scanned picture information to the server in a wireless manner.

3) The server side recognizes the text information or chemical symbols in the picture through the OCR feature recognition module, that is, recognizes the text information "limestone" in the "limestone" picture.

4) For the text information of "limestone", use the search technology to perform full pattern matching against the cigarette material licensing standard database. If it is found that "limestone" appears on the license list, the result will be fed back to the user terminal for display to complete a match. However, in this embodiment, the cigarette material licensing standard database is shown in FIG. 5 , and “limestone” is not found in the licensing list, and the first matching fails, and a second matching is required. Among them, the information on the names of substances permitted in the licensing standards for smoking materials is stored in the smoking materials licensing standard database.

5) Operate on the feature set of the material name "limestone" text or the feature set of the picture segment, carry out fuzzy intelligent matching, and perform predictive matching in the historical experience database, as shown in Figure 6 for the historical experience database map. It was found that "limestone-calcium carbonate" is in the historical experience database, and the results are shown in Figure 7. The calculated similarity results are arranged from the top of the table. The closer to the top, the higher the similarity, indicating that "limestone" is used in cigarettes Within the scope of the material license, corresponding to the record No. 5 in the tobacco material license standard database, the result is displayed on the terminal, and the secondary matching process is completed.

6) According to the final matching result, the operator will confirm and automatically add the result to the historical experience database.

In addition, the present invention also provides a multi-process matching device for identifying substances used in smoking materials, including the following modules:

A module for scanning and identifying information on tobacco material in paper documents or pictures of tobacco material;

A module used to match the information of the identified smoking materials with the smoking materials licensing standard database, and if the matching is successful, the result will be displayed; wherein, the smoking materials licensing standard database stores the allowed smoking materials using information on substances;

A module used to match the information of tobacco materials with the historical experience database if the matching is unsuccessful, and display the results if the matching is successful; wherein, the historical experience database stores the cigarette materials that have been successfully or unsuccessfully matched before Information on substances used.

The above-mentioned multi-process matching device for identifying substances used in cigarette materials is actually a computer solution based on the corresponding method process of the present invention, that is, a software framework, and the above-mentioned various modules are each processing process or process corresponding to the method process. program. Since the introduction of the above method is sufficiently clear and complete, the device will not be described in detail.

Claims (8)

1. A method for identifying substances used by a multi-process matched tobacco material is characterized by comprising the following steps:

1) scanning and identifying information of the tobacco material in the paper file or the picture of the tobacco material;

2) matching the identified information of the tobacco material with a tobacco material permission standard database, and if the matching is successful, displaying the result; wherein the information of the allowable substances used by the tobacco materials is stored in the tobacco material permission standard database;

3) if the matching is unsuccessful, matching the information of the tobacco material with a historical experience library, and if the matching is successful, displaying the result; wherein the historical experience library stores information of the used substances of the cigarette materials which are successfully or unsuccessfully matched before;

the successful matching means that the information of the tobacco materials is determined to be stored in the tobacco material permission standard database or the historical experience database;

the database of the standard of permission for the tobacco material comprises calcium carbonate, and the historical experience database comprises the only mapping relation between limestone and calcium carbonate and CaCO₃Unique mapping relation with calcium carbonate and CaCO₃Unique mapping relation with calcium carbonate.

2. The method of claim 1, further comprising the step of storing the results of the matching with the historical experience repository in the historical experience repository.

3. The method for identifying substances used in multi-process matched tobacco materials according to claim 1 or 2, wherein when the information of the tobacco materials is matched with a historical experience base, a formula is adopted for a characteristic set of the information of the tobacco materials

Measuring, and displaying the result according to the matching similarity from high to low; wherein, C_iThe method comprises the steps of setting a central point of a feature set cluster, setting x as a set distribution point of the cluster, setting k as a feature data set number, and setting E as an optimal value of a result.

4. The method for identifying substances used by multi-process matched tobacco materials as claimed in claim 1 or 2, wherein the information of the tobacco materials in the paper documents or pictures of the tobacco materials is identified by OCR recognition.

5. The device for identifying the substances used by the tobacco materials in the multi-process matching manner is characterized by comprising the following modules:

a module for scanning and identifying information of the tobacco material in the paper document or picture of the tobacco material;

a module for matching the identified information of the tobacco material with a database of allowable standard for tobacco material, and displaying the result if the matching is successful; wherein the information of the allowable substances used by the tobacco materials is stored in the tobacco material permission standard database;

a module for matching the information of the tobacco material with the historical experience base if the matching is unsuccessful, and displaying the result if the matching is successful; wherein the historical experience library stores information of the used substances of the cigarette materials which are successfully or unsuccessfully matched before;

the successful matching means that the information of the tobacco materials is determined to be stored in the tobacco material permission standard database or the historical experience database;

the database of the standard of permission for the tobacco material comprises calcium carbonate, and the historical experience database comprises the only mapping relation between limestone and calcium carbonate and CaCO₃Unique mapping relation with calcium carbonate and CaCO₃Unique mapping relation with calcium carbonate.

6. The apparatus for identifying substances for use with multiple processes matched with smoking materials of claim 5, further comprising a module for storing the results of the matching with the historical experience library in the historical experience library.

7. The device for identifying substances used in multi-process matched tobacco materials as claimed in claim 5 or 6, wherein when the information of the tobacco materials is matched with the historical experience base, a formula is adopted for the characteristic set of the information of the tobacco materials

Measuring, and displaying the result according to the matching similarity from high to low; wherein, C_iThe method comprises the steps of setting a central point of a feature set cluster, setting x as a set distribution point of the cluster, setting k as a feature data set number, and setting E as an optimal value of a result.

8. The device for identifying the substance used by the multi-process matched tobacco material as claimed in claim 5 or 6, wherein the information of the tobacco material in the paper file or the picture of the tobacco material is identified by OCR recognition.

Images