CN115661839A - Bill filing method, device, terminal and storage medium - Google Patents

Abstract

The invention relates to the technical field of bill word processing, in particular to a bill archiving method, a device, a terminal and a storage medium, wherein the method firstly acquires a bill image; then, identifying the bill image to obtain a plurality of character sets; then, acquiring a plurality of word collections according to the word stock and the plurality of character collections; and finally, filling the plurality of word collections into a plurality of item tables according to the relevance of the bills and the business processes. According to the embodiment of the invention, the bill images are separated to obtain the images of the plurality of bills, and the characters of the plurality of bills are identified based on the images of the plurality of bills, so that the bills can be operated in batch, and the processing efficiency is improved. And then, performing word segmentation according to the characters and word stock of the bill to obtain an index of the word segmentation, and filling the completed item table according to the index of the word segmentation, wherein the filling accuracy is high because the filling is performed according to the relevance of the index.

Description

Bill archiving method, device, terminal and storage medium

technical field

The invention relates to the technical field of bill word processing, and in particular to a bill filing method, device, terminal and storage medium.

Background technique

At present, in order to reduce the bill review and storage work of financial personnel, and save manpower and material resources in this regard, existing technical personnel have created automated bill storage machines, bill filing machines, etc., such as a prior art provides a bill The automatic filing system realizes the automatic storage of bills through the bill image forming mechanism and the bill paper filing and storage device.

However, when filing bills, users often store multiple bills at one time for high-speed scanning to digitize data. When the high-speed scanning device is working, batches of bills are put into the scanning device, and the high-speed scanning device performs high-speed scanning, that is, batch Loaded paper is fed at a high speed, fast scanning is performed and the scan results (i.e. scanned images) are buffered, and batches of paper are discharged at high speed. This process does not involve identifying the pages contained in the image to distinguish the individual copies of the document. A bill includes a cover page and an invoice page. The cover page is a paper with information including but not limited to the invoice details, amount, taxpayer identification number and/or signature of the applicant (reviewer), and the invoice page is the paper on which the invoice is pasted. paper.

If each bill is recognized while scanning, there will be the following problems:

1. The scanning results of each page must be judged in real time, which will cause the recognition process to be triggered for each page, resulting in the interruption of high-speed scanning, so that high-speed scanning equipment cannot take advantage of its batch high-speed scanning characteristics and affect the scanning speed;

2. During the identification process, only by identifying the first page of the next copy can we know the completion of the last copy, and for the scanner, it cannot return the scanned paper, so that the bill will be mistakenly discharged when divided The first page of the next copy is given to the previous note.

Based on this, it is necessary to develop and design a bill filing method.

Contents of the invention

The embodiments of the present invention provide a bill filing method, device, terminal and storage medium, which are used to solve the problem in the prior art that bills need to be processed one by one when filing, and the efficiency is relatively low.

In the first aspect, the embodiment of the present invention provides a bill filing method, including:

Obtaining a bill image, wherein the bill image includes a plurality of image blocks, and each image block contains an image of a bill;

Recognizing the bill image, and acquiring a plurality of character sets, wherein the plurality of character sets correspond to a plurality of bills, and the character sets include face characters of the bill;

Obtaining multiple vocabulary sets according to the lexicon and the multiple character sets, wherein the multiple vocabulary sets correspond to the multiple bills, and the vocabulary set includes multiple vocabulary representing the face text of the bill;

Filling the plurality of vocabulary sets into a plurality of item tables according to the association between the bill and the business process, wherein the business process is associated with the plurality of item tables.

In a possible implementation manner, before the recognition of the bill image and the acquisition of multiple character sets, the steps include:

Decolorizing the bill image to obtain a grayscale image;

Adjusting the resolution of the grayscale image to obtain an image with a preset resolution, wherein the adjustment of the resolution includes at least one of the following: compression, cropping, and pixel expansion through difference;

Adjusting the multiple pixels of the grayscale image so that the multiple pixels of the grayscale image are distributed in a preset interval, wherein the preset interval represents the overall grayscale interval of the grayscale image;

The grayscale curves of the pixels of the grayscale image are adjusted to conform to the preset grayscale distribution.

In a possible implementation manner, the adjusting the multiple pixels of the grayscale image so that the multiple pixels of the grayscale image are distributed in a preset interval includes:

Acquiring multiple grayscale values corresponding to multiple pixels in the grayscale image;

Gray value adjustment step: adjusting the plurality of gray values according to a first formula, wherein the first formula is:

In the formula, p _id is the adjusted gray value of the i-th interval in the gray-scale image, p _io is the gray-scale value of the i-th interval in the gray-scale image before adjustment, and p _min is the gray-scale value before the interval adjustment in the gray-scale image The minimum value, p _max is the maximum value before the interval adjustment in the grayscale image, a is the maximum value of the preset interval, and b is the minimum value of the preset interval.

In a possible implementation manner, the adjusting the grayscale curves of the plurality of pixels of the grayscale image to conform to a preset grayscale distribution includes:

Acquiring multiple grayscale values corresponding to multiple pixels in the grayscale image;

Sorting the plurality of grayscale values according to the size of the value, and dividing them into a plurality of grayscale sets according to preset intervals of the plurality of values;

Counting the number of grayscale values in multiple grayscale sets to obtain a grayscale interval sequence, wherein each element in the grayscale interval sequence represents the number of grayscale values in the interval of the corresponding value;

If the deviation between the gray value interval sequence and the sample sequence is greater than the threshold, adjust the plurality of gray values through the second formula according to the deviation between the gray value interval sequence and the sample sequence, and jump to the The gray value adjustment step, wherein the sample sequence represents the number of gray values in the interval corresponding to the gray image samples, and the second formula is:

In the formula, p _ic is the adjusted gray value of the i-th curve in the gray-scale image, p _id is the gray-scale value of the i-th curve in the gray-scale image before adjustment, c is the concave-convex coefficient, d is the position coefficient, b is the minimum value of the preset interval.

In a possible implementation manner, the identifying the bill image and acquiring multiple character sets includes:

extracting multiple image blocks according to the bill image, wherein the multiple image blocks correspond to multiple bills;

Send the multiple image blocks into the text recognition model respectively, and obtain multiple text sets corresponding to the multiple image blocks, wherein the text recognition model is based on the initial model constructed by the CNN neural network model, after Multiple training samples are obtained through training;

Wherein, the extracting a plurality of image blocks according to the bill image includes:

For the bill image, perform the following steps respectively from the horizontal and vertical directions:

Get the scan set, where the scan set is an empty set;

Value acquisition step: acquiring a plurality of pixel values of one row or one column from the starting position of the receipt image through the scan set;

calculating the sum of each element in the scan set as a scan sum;

If the scan sum is less than a threshold, mark the row or column where the scan set is located as blank;

If the scan set does not reach the end position of the bill image, move the position of the scan set in the bill image according to a preset moving distance, and jump to the step of obtaining a value;

Count the rows or columns of continuous blanks, and use the rows or columns of the continuous blanks exceeding the threshold as the cutting line;

Cutting the bill image according to multiple cutting lines to obtain multiple image blocks.

In a possible implementation manner, the multiple vocabulary sets are acquired according to the thesaurus and the multiple word sets, wherein the multiple vocabulary sets correspond to the multiple notes, and the vocabulary sets include multiple Vocabulary characterizing the face text of bills, including:

For each text set perform the following steps:

Obtain a pending index and a vocabulary set, the pending index is an empty index, and the vocabulary set is an empty vocabulary set;

From the beginning position of the text set, select a word as the undetermined word;

Searching step: searching for the undetermined word from the thesaurus;

If the undetermined word is included in the thesaurus, the obtained index of the undetermined word is used as the undetermined index, and a word after the undetermined word is selected, added to the undetermined word, and jumps to the Describe the search steps;

If the undetermined word is not included in the thesaurus, and the undetermined index is an empty index, then select a character after the undetermined word, add it to the undetermined word, and jump to the search step ;

If the undetermined word is not included in the thesaurus, and the undetermined index is a non-empty index, then the undetermined index is added to the vocabulary set, and the undetermined index is set as an empty index. If the undetermined word does not reach the end of the word set, select a word after the undetermined word as the undetermined word, and jump to the search step.

In a possible implementation manner, according to the correlation between the bill and the business process, the multiple vocabulary sets are filled into multiple item tables, including:

Obtaining a business flow and a plurality of item tables related to the business flow, wherein the business flow is associated with the plurality of notes;

For each project table, perform the following steps:

Extracting list items and items to be filled, wherein, the items to be filled are filled with vocabulary according to the list items;

Find the index of the list item from the thesaurus according to the list item;

Find the vocabulary with the smallest index difference with the list item from the vocabulary set, and fill in the item corresponding to the list item, wherein the corresponding vocabulary in the vocabulary set is arranged adjacently.

In the second aspect, the embodiment of the present invention provides a bill filing device, which is used to implement the bill filing method described in the first aspect or any possible implementation manner of the first aspect, the bill filing device includes:

An image acquisition module, configured to acquire a bill image, wherein the bill image includes a plurality of image blocks, and each image block contains an image of a bill;

A character recognition module, configured to recognize the bill image and acquire multiple character sets, wherein the multiple character sets correspond to multiple bills, and the character sets include the face characters of the bills;

A word segmentation module, configured to acquire a plurality of vocabulary sets according to the thesaurus and the plurality of character sets, wherein the plurality of vocabulary sets correspond to the plurality of notes, and the vocabulary set includes a plurality of words representing the face text of the note ;

as well as,

The form filling module is configured to fill the plurality of vocabulary sets into a plurality of item tables according to the association between the bill and the business process, wherein the business process is associated with the plurality of item tables.

In the third aspect, the embodiment of the present invention provides a terminal, including a memory and a processor, the memory stores a computer program that can run on the processor, and the processor implements the above when executing the computer program Steps in the method described in the first aspect or any possible implementation manner of the first aspect.

In a fourth aspect, the embodiment of the present invention provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, any of the above first aspect or the first aspect is implemented. A possible implementation manner is the steps of the described method.

The beneficial effect that the embodiment of the present invention exists compared with prior art is:

A bill filing method disclosed in the embodiment of the present invention, which first acquires a bill image, wherein the bill image includes a plurality of image blocks, and each image block contains an image of a bill; then, the bill image is processed Identifying and obtaining multiple character sets, wherein the multiple character sets correspond to multiple bills, and the character sets include the face characters of the bills; then, according to the lexicon and the multiple character sets, obtain multiple vocabulary sets, wherein, The plurality of vocabulary sets correspond to the plurality of bills, and the vocabulary set includes a plurality of vocabulary representing the face text of the bills; finally, according to the relevance between the bill and the business process, the plurality of vocabulary sets are filled in a plurality of items table, wherein the business process is associated with the plurality of project tables. In the embodiment of the present invention, the bill images are separated, multiple bill images are obtained, and the characters of multiple bills are recognized based on the multiple bill images. Therefore, the bills can be operated in batches, and the processing efficiency is improved. Then, according to the text and thesaurus of the bill, perform word segmentation to obtain the index of the word segmentation. According to the index of the word segmentation, the filling of the completed item table is based on the relevance of the index, so the accuracy of filling is high.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings that need to be used in the embodiments or descriptions of the prior art. Obviously, the drawings in the following descriptions are only of the present invention. For some embodiments, those of ordinary skill in the art can also obtain other drawings based on these drawings without paying creative efforts.

Fig. 1 is a flow chart of the bill filing method provided by the embodiment of the present invention;

Fig. 2 is a schematic diagram of an image separation process provided by an embodiment of the present invention;

Fig. 3 is a functional block diagram of the bill filing device provided by the embodiment of the present invention;

Fig. 4 is a functional block diagram of a terminal provided by an embodiment of the present invention.

Detailed ways

In the following description, specific details such as specific system structures and techniques are presented for illustration rather than limitation, so as to thoroughly understand the embodiments of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, devices, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.

In order to make the purpose, technical solution and advantages of the present invention more clear, the following will describe through specific implementation manners in conjunction with the accompanying drawings.

Below the embodiment of the present invention is described in detail, and this example is carried out under the premise of technical solution of the present invention, has provided detailed implementation mode and concrete operation process, but protection scope of the present invention is not limited to following embodiment .

Fig. 1 is a flow chart of a bill filing method provided by an embodiment of the present invention.

As shown in Figure 1, it shows the implementation flow chart of the bill filing method provided by the embodiment of the present invention, which is described in detail as follows:

In step 101, a bill image is acquired, wherein the bill image includes a plurality of image blocks, and each image block contains an image of a bill.

In step 102, the image of the bill is recognized, and multiple character sets are acquired, wherein the multiple character sets correspond to multiple bills, and the character sets include the characters on the face of the bill.

In some embodiments, the step 102 includes:

extracting multiple image blocks according to the bill image, wherein the multiple image blocks correspond to multiple bills;

Send the multiple image blocks into the text recognition model respectively, and obtain multiple text sets corresponding to the multiple image blocks, wherein the text recognition model is based on the initial model constructed by the CNN neural network model, after Multiple training samples are obtained through training;

Wherein, the extracting a plurality of image blocks according to the bill image includes:

For the bill image, perform the following steps respectively from the horizontal and vertical directions:

Get the scan set, where the scan set is an empty set;

Value acquisition step: acquiring a plurality of pixel values of one row or one column from the starting position of the receipt image through the scan set;

calculating the sum of each element in the scan set as a scan sum;

If the scan sum is less than a threshold, mark the row or column where the scan set is located as blank;

If the scan set does not reach the end position of the bill image, move the position of the scan set in the bill image according to a preset moving distance, and jump to the step of obtaining a value;

Count the rows or columns of continuous blanks, and use the rows or columns of the continuous blanks exceeding the threshold as the cutting line;

Cutting the bill image according to multiple cutting lines to obtain multiple image blocks.

Exemplarily, in the embodiment of the present invention, the recognition of the bill image is based on dividing the image into multiple blocks, each block contains an image of a bill, and then inputting the image into the recognition model for recognition.

The recognition model is obtained based on the initial model training, and in some application scenarios, the initial model is constructed based on the CNN neural network model. Specifically, the CNN neural network model includes a convolutional layer, a pooling layer, and a fully connected hidden layer. Layer and output layer, the convolutional layer is used to extract the main features of the image through convolution, and the pooling layer is used to further collect features and reduce the amount of feature data. The data output by the pooling layer is fully connected hidden After the inclusion layer is calculated, the text included in the indicated image is output through the output layer.

For the training step, specifically, it is to send image samples of multiple characters into the initial model, and adjust the parameters of the model through the deviation between the output result of the initial model and the text, that is, to gradually improve the initial model through iteration. recognition accuracy until the model has the expected recognition accuracy.

For the image separation step, the basic principle is to identify the boundaries of multiple bill images. As shown in FIG. The blank between the block line spacing, by identifying the blank width, you can find the separation line and separate the image.

In one application scenario, the segmentation is to scan the set 204 in a predetermined order, for example, from top to bottom, from left to right horizontally or vertically to take out the pixel values of one row or one column in the image. In FIG. 2, it is from From top to bottom, take out the pixel value where the horizontal line 203 is located, and put it into the scan set 204 .

After scanning, the sum of each element in the scan set 204 is calculated. If the sum is less than the threshold value, it indicates that the row is blank, and the row is marked as blank. Then, the horizontal row 203 is moved in a predetermined order, and the above steps are repeated.

If multiple consecutive lines are all blank, one of the multiple lines may be used as a cutting line, and cut according to the cutting line to obtain multiple image blocks.

Before step 102, including:

Decolorizing the bill image to obtain a grayscale image;

Adjusting the resolution of the grayscale image to obtain an image with a preset resolution, wherein the adjustment of the resolution includes at least one of the following: compression, cropping, and pixel expansion through difference;

Adjusting the multiple pixels of the grayscale image so that the multiple pixels of the grayscale image are distributed in a preset interval, wherein the preset interval represents the overall grayscale interval of the grayscale image;

The grayscale curves of the pixels of the grayscale image are adjusted to conform to the preset grayscale distribution.

In some implementation manners, the adjusting the multiple pixels of the grayscale image so that the multiple pixels of the grayscale image are distributed in a preset interval includes:

Acquiring multiple grayscale values corresponding to multiple pixels in the grayscale image;

Gray value adjustment step: adjusting the plurality of gray values according to a first formula, wherein the first formula is:

In the formula, p _id is the adjusted gray value of the i-th interval in the gray-scale image, p _io is the gray-scale value of the i-th interval in the gray-scale image before adjustment, and p _min is the gray-scale value before the interval adjustment in the gray-scale image The minimum value, p _max is the maximum value before the interval adjustment in the grayscale image, a is the maximum value of the preset interval, and b is the minimum value of the preset interval.

In some implementation manners, the adjusting the grayscale curves of the plurality of pixels of the grayscale image to conform to a preset grayscale distribution includes:

Acquiring multiple grayscale values corresponding to multiple pixels in the grayscale image;

Sorting the plurality of grayscale values according to the size of the value, and dividing them into a plurality of grayscale sets according to preset intervals of the plurality of values;

Counting the number of grayscale values in multiple grayscale sets to obtain a grayscale interval sequence, wherein each element in the grayscale interval sequence represents the number of grayscale values in the interval of the corresponding value;

If the deviation between the gray value interval sequence and the sample sequence is greater than the threshold, adjust the plurality of gray values through the second formula according to the deviation between the gray value interval sequence and the sample sequence, and jump to the The gray value adjustment step, wherein the sample sequence represents the number of gray values in the interval corresponding to the gray image samples, and the second formula is:

In the formula, p _ic is the adjusted gray value of the i-th curve in the gray-scale image, p _id is the gray-scale value of the i-th curve in the gray-scale image before adjustment, c is the concave-convex coefficient, d is the position coefficient, b is the minimum value of the preset interval.

Exemplarily, as we know, images maintain good consistency, which is beneficial to the process of image recognition, so it is essential for preprocessing before recognition.

In the embodiment of the present invention, the preprocessing of the image is to firstly decolorize the image, then adjust the resolution of the image to a preset interval, then adjust the overall grayscale value interval of the image, and finally adjust each grayscale interval Distribution.

In terms of adjusting the overall grayscale range of the image, it uses the first formula to adjust the value of each pixel in the image:

In the formula, p _id is the adjusted gray value of the i-th interval in the gray-scale image, p _io is the gray-scale value of the i-th interval in the gray-scale image before adjustment, and p _min is the gray-scale value before the interval adjustment in the gray-scale image The minimum value, p _max is the maximum value before the interval adjustment in the grayscale image, a is the maximum value of the preset interval, and b is the minimum value of the preset interval.

After adjusting the pixel value interval of the overall image, it is necessary to adjust the grayscale curve of the image, that is, to adjust the distribution of each grayscale interval in the image.

First, count the number of pixel values in each grayscale interval. If the number of pixel values in each interval does not meet the expected requirements, for example, the proportion of grayscale intervals with low values is more, indicating that the grayscale curve is more convex. And if there are more grayscale intervals with high values, it means that the grayscale curve is more concave.

The bumpiness of the curve can be adjusted by applying the second formula:

In the formula, p _ic is the adjusted gray value of the i-th curve in the gray-scale image, p _id is the gray-scale value of the i-th curve in the gray-scale image before adjustment, c is the concave-convex coefficient, d is the position coefficient, b is the minimum value of the preset interval.

For a more concave curve, the concave-convex coefficient can be set to a negative value, and the gray-scale curve can be adjusted through the second formula to make its concavity smaller.

For a more convex curve, setting the concave-convex coefficient to a positive value can adjust the gray-scale curve through the second formula to make its convexity smaller.

The position coefficient is a coefficient to control the position of the bump. For example, if the bump is close to the rightmost position of the grayscale curve, set the position coefficient to a larger positive value. If the bump is close to the leftmost position of the grayscale curve , set the position factor to a small positive or negative value.

In step 103, a plurality of vocabulary sets are obtained according to the thesaurus and the plurality of character sets, wherein the plurality of vocabulary sets correspond to the plurality of bills, and the vocabulary sets include a plurality of vocabulary representing the face text of the bill .

In some implementations, step 103 includes:

For each text set perform the following steps:

Obtain a pending index and a vocabulary set, the pending index is an empty index, and the vocabulary set is an empty vocabulary set;

From the beginning position of the text set, select a word as the undetermined word;

Searching step: searching for the undetermined word from the thesaurus;

If the undetermined word is included in the thesaurus, the obtained index of the undetermined word is used as the undetermined index, and a word after the undetermined word is selected, added to the undetermined word, and jumps to the Describe the search steps;

If the undetermined word is not included in the thesaurus, and the undetermined index is an empty index, then select a character after the undetermined word, add it to the undetermined word, and jump to the search step ;

If the undetermined word is not included in the thesaurus, and the undetermined index is a non-empty index, then the undetermined index is added to the vocabulary set, and the undetermined index is set as an empty index. If the undetermined word does not reach the end of the word set, select a word after the undetermined word as the undetermined word, and jump to the search step.

Exemplarily, the thesaurus includes a plurality of words, and the plurality of words are arranged in order of correspondence, and the arrangement number is the index of the word. Correspondence refers to the answer pair of two words, for example, for the vocabulary of equipment name, the corresponding vocabulary includes transformer, disconnector, circuit breaker, router, switch and so on.

For the word segmentation of the text set, find the corresponding vocabulary in the thesaurus through text combination, which means that the word segmentation is successful. On this basis, if you add another word, then find the corresponding vocabulary in the thesaurus. If found, the newly added word will be added. As a new word, otherwise, the word participle is complete.

The index corresponding to the word segmentation is added to the vocabulary set, and the vocabulary set corresponding to the text set is obtained.

In step 104, the plurality of vocabulary sets are filled into a plurality of item tables according to the association between the ticket and the business process, wherein the business process is associated with the plurality of item tables.

In some embodiments, step 104 includes:

Obtaining a business flow and a plurality of item tables related to the business flow, wherein the business flow is associated with the plurality of notes;

For each project table, perform the following steps:

Extracting list items and items to be filled, wherein, the items to be filled are filled with vocabulary according to the list items;

Find the index of the list item from the thesaurus according to the list item;

Find the vocabulary with the smallest index difference with the list item from the vocabulary set, and fill in the item corresponding to the list item, wherein the corresponding vocabulary in the vocabulary set is arranged adjacently.

Exemplarily, generally speaking, one business flow corresponds to multiple project forms, and after associating multiple vocabulary sets with multiple project forms, the form filling process can be realized.

Specifically, the index of the list item is extracted based on the thesaurus, and the vocabulary with the smallest difference is found from the vocabulary set according to the index of the list item, and filled into the filling item.

The embodiment of the bill filing method of the present invention first acquires the bill image, wherein the bill image includes a plurality of image blocks, and each image block contains an image of a bill; then, the bill image is identified to obtain multiple A character set, wherein a plurality of character sets correspond to a plurality of bills, and the character set includes face characters of bills; then, according to the thesaurus and the plurality of character sets, a plurality of vocabulary sets are obtained, wherein the plurality of vocabulary The set corresponds to the plurality of bills, and the vocabulary set includes a plurality of vocabulary representing the face text of the bill; finally, according to the relevance between the bill and the business process, the plurality of vocabulary sets are filled in multiple item tables, wherein, The business process is associated with the multiple item tables. In the embodiment of the present invention, the bill images are separated, multiple bill images are obtained, and the characters of multiple bills are recognized based on the multiple bill images. Therefore, the bills can be operated in batches, and the processing efficiency is improved. Then, according to the text and thesaurus of the bill, perform word segmentation to obtain the index of the word segmentation. According to the index of the word segmentation, the filling of the completed item table is based on the relevance of the index, so the accuracy of filling is high.

It should be understood that the sequence numbers of the steps in the above embodiments do not mean the order of execution, and the execution order of each process should be determined by its functions and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

The following are device implementations of the present invention, and for details that are not described in detail, reference may be made to the corresponding method implementations above.

Fig. 3 is a functional block diagram of the bill filing device provided by the embodiment of the present invention. With reference to Fig. 3, the bill filing device 3 includes: an image acquisition module 301, a character recognition module 302, a word segmentation module 303 and a form filling module 304, wherein:

An image acquisition module 301, configured to acquire a receipt image, wherein the receipt image includes a plurality of image blocks, and each image block contains an image of a receipt;

A character recognition module 302, configured to recognize the bill image and acquire multiple character sets, wherein the multiple character sets correspond to multiple bills, and the character sets include the face characters of the bills;

The word segmentation module 303 is configured to acquire multiple vocabulary sets according to the thesaurus and the multiple character sets, wherein the multiple vocabulary sets correspond to the multiple bills, and the vocabulary sets include multiple character vocabulary;

The form filling module 304 is configured to fill the multiple vocabulary sets into multiple item tables according to the association between the bill and the business process, wherein the business process is associated with the multiple item tables.

Fig. 4 is a functional block diagram of a terminal provided by an embodiment of the present invention. As shown in FIG. 4 , the terminal 4 in this embodiment includes: a processor 400 and a memory 401 , and the memory 401 stores a computer program 402 that can run on the processor 400 . When the processor 400 executes the computer program 402, it realizes the steps in the above-mentioned various bill filing methods and implementations, for example, steps 101 to 104 shown in FIG. 1 .

Exemplarily, the computer program 402 may be divided into one or more modules/units, and the one or more modules/units are stored in the memory 401 and executed by the processor 400 to complete this invention.

The terminal 4 may be a computing device such as a desktop computer, a notebook, a palmtop computer, or a cloud server. The terminal 4 may include, but not limited to, a processor 400 and a memory 401 . Those skilled in the art can understand that FIG. 4 is only an example of the terminal 4 and does not constitute a limitation on the terminal 4. It may include more or less components than shown in the figure, or combine some components, or different components, such as The terminal 4 may also include an input and output device, a network access device, a bus, and the like.

The so-called processor 400 may be a central processing unit (Central Processing Unit, CPU), and may also be other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), Field-Programmable Gate Array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor, or the processor may be any conventional processor, or the like.

The storage 401 may be an internal storage unit of the terminal 4 , such as a hard disk or memory of the terminal 4 . The memory 401 may also be an external storage device of the terminal 4, such as a plug-in hard disk equipped on the terminal 4, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, Flash Card (Flash Card), etc. Further, the memory 401 may also include both an internal storage unit of the terminal 4 and an external storage device. The memory 401 is used to store the computer program 402 and other programs and data required by the terminal 4 . The memory 401 can also be used to temporarily store data that has been output or will be output.

Those skilled in the art can clearly understand that for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used for illustration. In practical applications, the above-mentioned functions can be assigned to different functional units, Completion of modules means that the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the implementation manner may be integrated into one processing unit, or each unit may exist separately physically, or two or more units may be integrated into one unit, and the above-mentioned integrated units may adopt hardware It can also be implemented in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the protection scope of the present application. For the specific working process of the units and modules in the above system, reference may be made to the corresponding process in the foregoing method implementation manners, and details are not repeated here.

In the foregoing implementation manners, the descriptions of each implementation manner have their own emphases, and for parts that are not detailed or recorded in a certain implementation manner, refer to the relevant descriptions of other implementation manners.

Those skilled in the art can appreciate that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementation should not be regarded as exceeding the scope of the present invention.

In the embodiments provided in the present invention, it should be understood that the disclosed device/terminal and method may be implemented in other ways. For example, the device/terminal implementation described above is only illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or Components may be combined or integrated into another system, or some features may be omitted, or not implemented. In another point, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

If the integrated module/unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention realizes all or part of the processes in the methods of the above-mentioned embodiments, and can also be completed by instructing related hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer When the program is executed by the processor, it can realize the steps of the above-mentioned methods and device implementations. Wherein, the computer program includes computer program code, and the computer program code may be in the form of source code, object code, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a removable hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (Read-Only Memory, ROM) , random access memory (Random Access Memory, RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still be applied to the foregoing embodiments Modifications to the technical solutions described in the method, or equivalent replacement of some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should be included in the within the protection scope of the present invention.

Claims (10)

1. A method for archiving documents, comprising:

acquiring a bill image, wherein the bill image comprises a plurality of image blocks, and each image block comprises an image of a bill;

identifying the bill image to obtain a plurality of character sets, wherein the character sets correspond to the bills and comprise the bill face characters of the bills;

acquiring a plurality of word collections according to a word bank and the character collections, wherein the word collections correspond to the notes and comprise a plurality of words representing the face characters of the notes;

and filling the plurality of word collections into a plurality of item tables according to the relevance between the bills and the business processes, wherein the business processes are relevant to the plurality of item tables.

2. The document filing method according to claim 1, wherein before said recognizing the document image and obtaining the plurality of character sets, comprising:

decolorizing the bill image to obtain a gray image;

adjusting the resolution of the gray image to obtain an image with a preset resolution, wherein the adjustment of the resolution comprises at least one of the following steps: compressing, clipping, and expanding pixels by difference;

adjusting a plurality of pixels of the gray-scale image so that the plurality of pixels of the gray-scale image are distributed in a preset interval, wherein the preset interval represents an interval of the integral gray scale of the gray-scale image;

and adjusting the gray curve of a plurality of pixels of the gray image to accord with preset gray distribution.

3. The bill filing method according to claim 2, wherein the adjusting the plurality of pixels of the grayscale image so that the plurality of pixels of the grayscale image are distributed in a preset interval comprises:

acquiring a plurality of gray values corresponding to a plurality of pixels in the gray image;

gray value adjustment: adjusting the plurality of gray values according to a first formula, wherein the first formula is as follows:

in the formula, p _id Adjusted gray value, p, for the ith interval in a gray image _io For the gray value before the i-th interval adjustment in the gray image, p _min For the minimum value before the spatial adjustment in the grayscale image, o _max The maximum value before the interval adjustment in the gray-scale image is shown, wherein a is the maximum value of the preset interval, and b is the minimum value of the preset interval.

4. The document filing method according to claim 3, wherein the adjusting the gray curve of the plurality of pixels of the gray image to conform to a preset gray distribution comprises:

acquiring a plurality of gray values corresponding to a plurality of pixels in the gray image;

sorting the plurality of gray values according to the size of the values, and dividing the gray values into a plurality of gray sets according to preset intervals of a plurality of values;

counting the number of gray values in a plurality of gray level sets to obtain a gray level interval sequence, wherein each element in the gray level interval sequence represents the number of gray values in an interval of corresponding values;

if the deviation between the gray value interval sequence and the sample sequence is greater than a threshold value, adjusting the plurality of gray values according to the deviation between the gray value interval sequence and the sample sequence through a second formula, and skipping to the gray value adjusting step, wherein the sample sequence represents the number of gray values in an interval of corresponding values of a gray map sample, and the second formula is as follows:

in the formula, p _ic Adjusted gray value, p, for the ith curve in the gray image _id The gray value before the ith curve in the gray image is adjusted, c is a concave-convex coefficient, d is a position coefficient, and b is the minimum value of a preset interval.

5. The document archiving method according to claim 1, wherein said recognizing the document image to obtain a plurality of text sets comprises:

extracting a plurality of image blocks according to the bill images, wherein the plurality of image blocks correspond to the plurality of bills;

respectively sending the image blocks into a character recognition model, and acquiring a plurality of character sets corresponding to the image blocks, wherein the character recognition model is obtained by training a plurality of training samples based on an initial model constructed by a CNN neural network model;

wherein, the extracting a plurality of image blocks according to the bill image comprises:

respectively executing the following steps from the horizontal direction and the vertical direction for the bill image respectively:

acquiring a scanning set, wherein the scanning set is an empty set;

a value taking step: acquiring a plurality of pixel values of one row or one column from the starting position of the bill image through the scanning set;

calculating the sum of each element in the scanning set as a scanning sum;

if the scanning sum is smaller than the threshold value, marking the row or column where the scanning set is located as blank;

if the scanning set does not reach the tail end position of the bill image, moving the position of the scanning set in the bill image according to a preset moving distance, and skipping to the value taking step;

counting rows or columns of continuous blanks, and taking the rows or columns of the continuous blanks exceeding a threshold value as cutting lines;

and cutting the bill image according to the plurality of cutting lines to obtain a plurality of image blocks.

6. The document filing method according to any one of claims 1 to 5, wherein said obtaining a plurality of vocabulary sets from a lexicon and said plurality of character sets comprises:

for each set of words the following steps are performed:

acquiring an undetermined index and a word set, wherein the undetermined index is a null index, and the word set is a null word set;

selecting a character from the starting position of the character set as a word to be determined;

a searching step: searching the word to be determined from the word bank;

if the word bank comprises the undetermined word, taking the acquired index of the undetermined word as an undetermined index, selecting a word behind the undetermined word, adding the word into the undetermined word, and jumping to the searching step;

if the word bank does not contain the undetermined word and the undetermined index is a null index, selecting a word behind the undetermined word, adding the word into the undetermined word, and jumping to the searching step;

if the word bank does not contain the undetermined word and the undetermined index is a non-empty index, adding the undetermined index into the word set, setting the undetermined index as an empty index, and if the undetermined word does not reach the tail end of the word set, selecting a word behind the undetermined word as the undetermined word and jumping to the searching step.

7. The method of claim 6, wherein populating the plurality of vocabulary entries into a plurality of item tables based on the association of the ticket with the business process comprises:

acquiring a service flow and a plurality of item tables related to the service flow, wherein the service flow is associated with the plurality of tickets;

for each item table, the following steps are performed:

extracting a list item and an item to be filled, wherein the item to be filled fills a vocabulary according to the list item;

according to the list item, finding the index of the list item from the word stock;

and finding out the vocabulary with the minimum index difference value with the list item from the vocabulary set, and filling the filling item corresponding to the list item, wherein the vocabularies corresponding to the vocabulary set are arranged adjacently.

8. A bill filing apparatus for realizing the bill filing method according to any one of claims 1 to 7, the bill filing apparatus comprising:

the image acquisition module is used for acquiring a bill image, wherein the bill image comprises a plurality of image blocks, and each image block comprises an image of a bill;

the character recognition module is used for recognizing the bill images to obtain a plurality of character sets, wherein the character sets correspond to the bills, and the character sets comprise the bill face characters of the bills;

the word segmentation module is used for acquiring a plurality of word collections according to a word bank and the character collections, wherein the word collections correspond to the bills, and comprise a plurality of vocabularies representing the characters on the bill surfaces;

and the number of the first and second groups,

and the table filling module is used for filling the plurality of word collections into a plurality of project tables according to the relevance between the bills and the business process, wherein the business process is relevant to the plurality of project tables.

9. A terminal comprising a memory and a processor, the memory having stored therein a computer program operable on the processor, wherein the processor, when executing the computer program, performs the steps of the method according to any of claims 1 to 7.

10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.

Images