CN110209632A - A kind of electronics folder with case production, turn shelves system - Google Patents
A kind of electronics folder with case production, turn shelves system Download PDFInfo
- Publication number
- CN110209632A CN110209632A CN201910448250.4A CN201910448250A CN110209632A CN 110209632 A CN110209632 A CN 110209632A CN 201910448250 A CN201910448250 A CN 201910448250A CN 110209632 A CN110209632 A CN 110209632A
- Authority
- CN
- China
- Prior art keywords
- image
- training
- electronics folder
- module
- electronics
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/10—File systems; File servers
- G06F16/11—File system administration, e.g. details of archiving or snapshots
- G06F16/113—Details of archiving
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/18—Legal services; Handling legal documents
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/41—Analysis of document content
- G06V30/413—Classification of content, e.g. text, photographs or tables
Abstract
The present invention provides a kind of electronics folders with case production, turns shelves system, including acquisition module, and memory module receives the image for being uploaded to processing server, forms storage catalogue according to the gomma uploaded；Processing module replicates the image received from storage catalogue, is rectified a deviation to the image of duplication, removes black surround, removes noise, removal blank page and be inverted rotation processing；Categorization module, by processing module, treated that image detects, meet the content of text of archiving requirements obtained in picture and text by OCR, respective file title and page number information are extracted according to text formatting, judge file association, and the training pattern unit according to setting is repeatedly screened and Corresponding matching rule carries out document classification；Electronics folder produces module, generates electronics folder according to classification results, turns shelves module, takes distribution storage and cloud backup storage to carry out turning shelves in the electronics folder of generation.
Description
Technical field
The present invention relates to a kind of electronics folder integration technology, especially a kind of electronics folder produces with case, turns shelves system.
Background technique
With the continuous propulsion of whole nation law court's information works at different levels, the intelligence of mark is turned to networking, sunlight, intelligence
Intelligent law court roots in courts across the country.In the information-based course of law court, the trial flow system based on structured message
Construction with the system of execution has been achieved for gratifying achievement.Law court's data manage platform concentratedly and gradually manage kimonos to big data
Business platform success transition and upgrade, executes, judicial administration and state treatment case study provide digitization and intelligence for trial
The means of services.
Undoubtedly, informationization is to realize that people's court's service facilitation, trial is intelligent, executes high efficiency, Administration and Registration Division
Huas, the foundation stone for disclosing normalization, decision precision.
And reality is, the construction object of early stage information system is substantially to manage.Therefore for many judges
For, use information system though a compulsory task, not only information system without working Reduction of Students' Study Load for beach trial, instead
As a kind of burden and burden.
To find out its cause, about 90% is unstructured data, as party and trustee mention in the data source of law court
Various types of materials, document, audio-video of friendship etc..And the structured message of practical trial system on the court, it is defined in method mark file
Many fields, these materials can be derived from.But due to the limitation of technology and cognition, the value of this some materials data is still
It is not mined utilization, is equivalent to the still a piece of wasteland without really opening up wasteland.Wherein, many law courts be just can be phase after winding up the case
It closes material electronicsization and mounts system, this is the especially typical fact.
With implementing in full for registration system of putting on record, the speed and efficiency put on record are all in double growth, but judge and the administration of justice
The number of support staff thus causes the situation of " the more people of case are few " there is no increasing considerably therewith.One judge and Si
Method support staff generally required to be responsible for the multiple cases of trial within the same period, while shouldering and drafting document, being sent to document
Deng the judicial back work of relevant and program-type to trial management, and to ensure that related folder material is really achieved and be scanned with case
With the requirement of generation.In this way, which the work of judge and judicial support staff are just once become with node by original only sweeping
Multiple-Scan and arrangement, workload are multiplied.
Summary of the invention
In order to solve the above-mentioned technical problem, the main object of the present invention be to provide a kind of electronics folder with case production, turn shelves
System.
Its technical solution is:
The present invention provides a kind of electronics folders with case production, turns shelves system, including
Acquisition module is uploaded to processing clothes by scanner scanning and by the local interface that uploads of Seamless integration- twain agreement
Business device,
Memory module receives the image for being uploaded to processing server, forms storage catalogue according to the gomma uploaded；
Processing module replicates the image received from storage catalogue, is rectified a deviation to the image of duplication, remove black surround, go noise, going
Except blank page and it is inverted rotation processing；
Categorization module, by processing module, treated that image detects, and meet archiving requirements obtains picture and text by OCR
Content of text in this extracts respective file title and page number information according to text formatting, judges file association, and foundation is set
Fixed training pattern unit is repeatedly screened and Corresponding matching rule carries out document classification；
Electronics folder produces module, generates electronics folder according to classification results,
Turn shelves module, take distribution storage and cloud backup storage to carry out turning shelves in the electronics folder of generation,
The training pattern unit carries out modelling training to electronics folder and remembers and by force by collecting the electronics folder formed
Change training rules, and memory and intensive training rule are used for categorization module and carry out document classification.
Further, the correction includes the following steps:
The effective foreground area of image is obtained using ROI extractive technique,
Sub-zone dividing is carried out to effective foreground area, and demarcates and divides format and parameter,
Directional feature extraction is carried out to each all subregion,
It is measured according to direction character detection all subregion inclination,
According to format and parameter progress reverse process are divided, all subregion measurement results are merged, the integral inclined measurement of image is obtained,
Oblique correction is cut in completion.
Further, described that black surround is gone to include the following steps:
The gamma characteristic of detection image obtains the grey level histogram of image,
Grey level histogram and normalization histogram are calculated,
Image grayscale mean value is calculated according to normalization histogram,
The zeroth order W [i] and a class interval U [i] of normalization histogram are obtained according to image grayscale mean value,
It is handled by above-mentioned class interval U [i], obtains the corresponding gray value MAX of inter-class variance MAX,
Image segmentation is obtained with this,
Based on black surround physical location, applied morphology operation is real to black surround region detection,
Integrated location, area attributive character remove pseudo- black surround region, realize area filling, and application filtering to remaining black surround region
Technology realizes black surround edge-smoothing, and image segmentation is connected to domain lookup with morphology and effectively realizes doubtful black surround zone location, removes
The effective information retention of pseudo- black surround regional support.
Further, described that noise is gone to realize noise removal by application image enhancing and image filtering technology.
Further, the removal blank page includes completing image denoising using noise removal technology, is based on image histogram
Map analysis exports Image blank degree, returns to blank extent index.
Further, the matching rule is directory name belonging to file title is corresponding.
Further, it is as follows to carry out the trained step of modelling for the training pattern unit:
Training sample is collected according to the electronics folder of generation, and training sample is demarcated；
The training sample of calibration is at least constructed into one group of training dataset,
The text image of each group of training dataset of building is pre-processed, image is cut, image comparison degree is returned
One change processing,
Normalized data text image, is handled, output category result by convolutional neural networks structure,
The classification results of output and true tag along sort are compared, loss module is established, calculates loss,
According to the outcome evaluation classification results for calculating loss, according to training data and test data classification analysis network model, according to
According to the accuracy that classification analysis network model is classified, classification is completed.
Further, the convolutional neural networks structure includes convolutional layer, pond layer, full articulamentum and Softmax layers.
Further, the training pattern unit carry out modelling training can also be by cyclic convolution neural network structure
It is trained, step includes
The whole of input data or certain a part, handled using convolutional neural networks structure, output treated as a result,
By treated, result is compared with true information, is input to convolutional neural networks again according to the method for BP backpropagation
Structure is handled, export it is handling again as a result, will treated again that result is compared with true information, according to BP
The method circular treatment of backpropagation.
Further, the cycle-index that the cyclic convolution neural network structure is trained is not less than 2 times.
The present invention to case folder by, from automatic collection, image optimization processing, forming electronics folder and automatically generating, help
Law court's unit promotes electronics folder production rate, reduces the workload of judicial support staff, saves the scanning expense of electronic record.
Judicial support staff only needs simply to input Reference Number, papery folder is put into scanner, a key scanning is automatic to upload.
Ensure the sequence, in the right direction when scanning folder, neatly, the overall time of image procossing optimization will be directly influenced, make to generate
The efficiency of electronics folder improves, and has also confirmed the importance of regularized operation simultaneously, and the two complements each other.Passing judge and auxiliary
The operation system that personnel use scans a files, needs to carry out 8-9 operation in system formally start to scan, in addition volume
Ancestor uploads the time of classification, and producing efficiency is severely impacted.Present system provides the function that a key generates, and realizes not addendum
Add people, do not tie up fund, do not work extra shifts or extra hours, relies only on existing judicial support staff troop and a small amount of fund, just solve clothes
Business outsourcing or full-time mechanic mode need the problem of spending millions of members and a large amount of manpower and material resources working times just to can solve, conscientiously
It promotes case and handles efficiency.
The image file for scanning system can be according to the recognition result of OCR, the characteristics of image of scanning file and two dimensional code
Decoding result judges document type, the every lawsuit material of typing can automatic screening classification, be articulated to corresponding text
In bibliography record.
Recognition with Recurrent Neural Network can play the training of NextState by the information of laststate when being directed to text-processing
Certain booster action.By the way that in text, the collected weight got in network training before of the text of lastrow is also made
It goes to carry out neural metwork training for input information, and the key message of the pixel of all rows before text saves and assists net
Network training is to reach pinpoint accuracy.
Detailed description of the invention
Fig. 1 is flow chart of the invention；
Fig. 2, Fig. 3 are Filtered Embodiment figure in the present invention；
Fig. 4 is the flow chart that training pattern unit carries out modelling training in the present invention；
Fig. 5 is the flow chart that cyclic convolution neural network structure is trained in the present invention.
Specific embodiment
Below in conjunction with attached drawing and specific embodiment, the present invention will be described in detail, herein illustrative examples of the invention
And explanation is used to explain the present invention, but not as a limitation of the invention.
Referring to Fig.1, the present invention provides the present invention provides a kind of electronics folder with case production, turn shelves system, including
Acquisition module is uploaded to processing clothes by scanner scanning and by the local interface that uploads of Seamless integration- twain agreement
Business device,
Memory module receives the image for being uploaded to processing server, forms storage catalogue according to the gomma uploaded；
Processing module replicates the image received from storage catalogue, is rectified a deviation to the image of duplication, remove black surround, go noise, going
Except blank page and it is inverted rotation processing；
Categorization module, by processing module, treated that image detects, and meet archiving requirements obtains picture and text by OCR
Content of text in this extracts respective file title and page number information according to text formatting, judges file association, and foundation is set
Fixed training pattern unit is repeatedly screened and Corresponding matching rule carries out document classification；
Electronics folder produces module, generates electronics folder according to classification results,
Turn shelves module, take distribution storage and cloud backup storage to carry out turning shelves in the electronics folder of generation,
The training pattern unit carries out modelling training to electronics folder and remembers and by force by collecting the electronics folder formed
Change training rules, and memory and intensive training rule are used for categorization module and carry out document classification.
Further, the correction includes the following steps:
The effective foreground area of image is obtained using ROI extractive technique,
Sub-zone dividing is carried out to effective foreground area, and demarcates and divides format and parameter,
Directional feature extraction is carried out to each all subregion,
It is measured according to direction character detection all subregion inclination,
According to format and parameter progress reverse process are divided, all subregion measurement results are merged, the integral inclined measurement of image is obtained,
Oblique correction is cut in completion.
Further, described that black surround is gone to include the following steps:
The gamma characteristic of detection image obtains the grey level histogram of image,
Grey level histogram and normalization histogram are calculated,
Image grayscale mean value is calculated according to normalization histogram,
The zeroth order W [i] and a class interval U [i] of normalization histogram are obtained according to image grayscale mean value,
It is handled by above-mentioned class interval U [i], obtains the corresponding gray value MAX of inter-class variance MAX,
Image segmentation is obtained with this,
Based on black surround physical location, applied morphology operation is real to black surround region detection,
Integrated location, area attributive character remove pseudo- black surround region, realize area filling, and application filtering to remaining black surround region
Technology realizes black surround edge-smoothing, and image segmentation is connected to domain lookup with morphology and effectively realizes doubtful black surround zone location, removes
The effective information retention of pseudo- black surround regional support.
Further, described that noise is gone to realize noise removal by application image enhancing and image filtering technology.
Further, the removal blank page includes completing image denoising using noise removal technology, is based on image histogram
Map analysis exports Image blank degree, returns to blank extent index.
Further, the matching rule is directory name belonging to file title is corresponding.
Referring to Fig. 4, the step that the training pattern unit carries out modelling training is as follows:
Training sample is collected according to the electronics folder of generation, and training sample is demarcated；
The training sample of calibration is at least constructed into one group of training dataset,
The text image of each group of training dataset of building is pre-processed, image is cut, image comparison degree is returned
One change processing,
Normalized data text image, is handled, output category result by convolutional neural networks structure,
The classification results of output and true tag along sort are compared, loss module is established, calculates loss,
According to the outcome evaluation classification results for calculating loss, according to training data and test data classification analysis network model, according to
According to the accuracy that classification analysis network model is classified, classification is completed.
Further, the convolutional neural networks structure includes convolutional layer, pond layer, full articulamentum and Softmax layers.
Referring to Fig. 5, the training pattern unit carries out modelling training can also be by cyclic convolution neural network structure
It is trained, step includes
The whole of input data or certain a part, handled using convolutional neural networks structure, output treated as a result,
By treated, result is compared with true information, is input to convolutional neural networks again according to the method for BP backpropagation
Structure is handled, export it is handling again as a result, will treated again that result is compared with true information, according to BP
The method circular treatment of backpropagation.
Further, the cycle-index that the cyclic convolution neural network structure is trained is not less than 2 times.
In the present invention, using the effective foreground area of ROI is extracted, sub-zone dividing merges with sub-district field result and figure is effectively reduced
Influence as noise, invalid image information to integral error correcting and position result improves correction performance.
In the present invention, by being based on black surround physical location to image segmentation, doubtful black surround area is realized in applied morphology operation
Domain detection, the attributive character such as integrated location, area remove pseudo- black surround region, realize area filling to remaining black surround region, and answer
Black surround edge-smoothing is realized with filtering technique.Image segmentation is connected to domain lookup with morphology and effectively realizes that doubtful black surround region is fixed
Position, removing pseudo- black surround regional support effective information retention, filtering technique keeps processing result more beautiful, and the whole black surround that promoted is gone
Except performance.
In the present invention, application image enhancing and image filtering technology realize noise removal, and image enhancement technique can be solved effectively
The problems such as certainly image improves picture contrast, can effectively solve image salt-pepper noise that may be present by filtering technique, reduces
Interference of the noise to subsequent detection improves the total quality and signal-to-noise ratio of image.
In the present invention, image denoising is completed using noise removal technology, Image blank is exported based on image histogram analysis
Degree returns to blank extent index, and noise removal effectively prevents noise image blank page missing inspection problem, and blank extent index makes
Blank page detection function is more flexible convenient.
The image file for scanning system can be according to the recognition result of OCR, the characteristics of image of scanning file and two dimensional code
Decoding result judges document type, then passes through neural network learning, the accuracy of Continuous optimization classification.Typing it is each
Item lawsuit material meeting automatic screening classification, is articulated in corresponding document catalogue.
Electronics folder with case it is synchronous generate by above-mentioned acquisition, upload, intelligent classification, auto-sequencing link after, can root
According to the classification results and ordering scenario of final folder, list in volume corresponding to the folder is automatically generated.
Electronics folder is detected after generating with case, is split automatically to it, and the image text of multiple JPG formats is become
Part, and its image resolution ratio is detected and is arranged automatically, to guarantee that the quality of image meets the requirement of electronic record.
In the present invention, the folder content of characteristics of image training is cut mainly for various certificates, receipt, proof of service network
The image file of the types such as figure and express delivery list.The reason is that (such as tagged word is red for the uncertain factor being likely to occur on image
Chapter or spot cover) it is more, judgement is only carried out to its classification merely by OCR and is less susceptible to.This partial document is by collecting mark
Note, establishes data set, image is repeatedly handled, and can just enter training pattern.Then, it when training pattern exports result, then needs
Manually the result of output is judged, and according to output as a result, the key parameter to training pattern is adjusted.So circulation
Whole process, continuous adjusting parameter, until model is to the classification accuracy of judgement of such image file.
Recognition with Recurrent Neural Network can play the training of NextState by the information of laststate when being directed to text-processing
Certain booster action.By the way that in text, the collected weight got in network training before of the text of lastrow is also made
It goes to carry out neural metwork training for input information, and the key message of the pixel of all rows before text saves and assists net
Network training is to reach pinpoint accuracy.
Technical solution disclosed in the embodiment of the present invention is described in detail above, specific implementation used herein
Example is expounded the principle and embodiment of the embodiment of the present invention, and the explanation of above embodiments is only applicable to help to understand
The principle of the embodiment of the present invention；At the same time, for those skilled in the art is being embodied according to an embodiment of the present invention
There will be changes in mode and application range, in conclusion the content of the present specification should not be construed as to limit of the invention
System.
Claims (10)
1. a kind of electronics folder produces with case, turns shelves system, which is characterized in that including
Acquisition module is uploaded to processing clothes by scanner scanning and by the local interface that uploads of Seamless integration- twain agreement
Business device,
Memory module receives the image for being uploaded to processing server, forms storage catalogue according to the gomma uploaded；
Processing module replicates the image received from storage catalogue, is rectified a deviation to the image of duplication, remove black surround, go noise, going
Except blank page and it is inverted rotation processing；
Categorization module, by processing module, treated that image detects, and meet archiving requirements obtains picture and text by OCR
Content of text in this extracts respective file title and page number information according to text formatting, judges file association, and foundation is set
Fixed training pattern unit is repeatedly screened and Corresponding matching rule carries out document classification；
Electronics folder produces module, generates electronics folder according to classification results,
Turn shelves module, take distribution storage and cloud backup storage to carry out turning shelves in the electronics folder of generation,
The training pattern unit carries out modelling training to electronics folder and remembers and by force by collecting the electronics folder formed
Change training rules, and memory and intensive training rule are used for categorization module and carry out document classification.
2. electronics folder according to claim 1 produces with case, turns shelves system, the correction includes the following steps:
The effective foreground area of image is obtained using ROI extractive technique,
Sub-zone dividing is carried out to effective foreground area, and demarcates and divides format and parameter,
Directional feature extraction is carried out to each all subregion,
It is measured according to direction character detection all subregion inclination,
According to format and parameter progress reverse process are divided, all subregion measurement results are merged, the integral inclined measurement of image is obtained,
Oblique correction is cut in completion.
3. electronics folder according to claim 1 produces with case, turns shelves system, which is characterized in that described to go the black surround to include
Following steps:
The gamma characteristic of detection image obtains the grey level histogram of image,
Grey level histogram and normalization histogram are calculated,
Image grayscale mean value is calculated according to normalization histogram,
The zeroth order W [i] and a class interval U [i] of normalization histogram are obtained according to image grayscale mean value,
It is handled by above-mentioned class interval U [i], obtains the corresponding gray value MAX of inter-class variance MAX,
Image segmentation is obtained with this,
Based on black surround physical location, applied morphology operation is real to black surround region detection,
Integrated location, area attributive character remove pseudo- black surround region, realize area filling, and application filtering to remaining black surround region
Technology realizes black surround edge-smoothing, and image segmentation is connected to domain lookup with morphology and effectively realizes doubtful black surround zone location, removes
The effective information retention of pseudo- black surround regional support.
4. electronics folder according to claim 1 produces with case, turns shelves system, which is characterized in that described that noise is gone to pass through
Application image enhancing and image filtering technology realize noise removal.
5. electronics folder according to claim 1 produces with case, turns shelves system, which is characterized in that the removal blank page
Image denoising is completed including application noise removal technology, Image blank degree is exported based on image histogram analysis, returns to blank
Extent index.
6. electronics folder according to claim 1 produces with case, turns shelves system, which is characterized in that the matching rule is
Directory name belonging to file title is corresponding.
7. electronics folder according to claim 1 produces with case, turns shelves system, which is characterized in that the training pattern list
The step that member carries out modelling training is as follows:
Training sample is collected according to the electronics folder of generation, and training sample is demarcated；
The training sample of calibration is at least constructed into one group of training dataset,
The text image of each group of training dataset of building is pre-processed, image is cut, image comparison degree is returned
One change processing,
Normalized data text image, is handled, output category result by convolutional neural networks structure,
The classification results of output and true tag along sort are compared, loss module is established, calculates loss,
According to the outcome evaluation classification results for calculating loss, according to training data and test data classification analysis network model, according to
According to the accuracy that classification analysis network model is classified, classification is completed.
8. electronics folder according to claim 7 produces with case, turns shelves system, which is characterized in that the convolutional Neural net
Network structure includes convolutional layer, pond layer, full articulamentum and Softmax layers.
9. electronics folder according to claim 1 produces with case, turns shelves system, which is characterized in that the training pattern list
Member, which carries out modelling training, to be trained by cyclic convolution neural network structure, and step includes
The whole of input data or certain a part, handled using convolutional neural networks structure, output treated as a result,
By treated, result is compared with true information, is input to convolutional neural networks again according to the method for BP backpropagation
Structure is handled, export it is handling again as a result, will treated again that result is compared with true information, according to BP
The method circular treatment of backpropagation.
10. electronics folder according to claim 1 produces with case, turns shelves system, which is characterized in that the cyclic convolution mind
The cycle-index being trained through network structure is not less than 2 times.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910448250.4A CN110209632A (en) | 2019-05-27 | 2019-05-27 | A kind of electronics folder with case production, turn shelves system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910448250.4A CN110209632A (en) | 2019-05-27 | 2019-05-27 | A kind of electronics folder with case production, turn shelves system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110209632A true CN110209632A (en) | 2019-09-06 |
Family
ID=67788892
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910448250.4A Pending CN110209632A (en) | 2019-05-27 | 2019-05-27 | A kind of electronics folder with case production, turn shelves system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110209632A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110597818A (en) * | 2019-09-19 | 2019-12-20 | 腾讯科技（深圳）有限公司 | Block chain based volume query method, device, equipment and storage medium |
| CN110675121A (en) * | 2019-09-23 | 2020-01-10 | 珠海市新德汇信息技术有限公司 | Method for collecting picture type file material |
| CN110675289A (en) * | 2019-09-29 | 2020-01-10 | 四川大学 | Method for compiling electronic file catalogue with case criminal review |
| CN110728240A (en) * | 2019-10-14 | 2020-01-24 | 北京华宇信息技术有限公司 | Method and device for automatically identifying title of electronic file |
| CN111445433A (en) * | 2019-10-14 | 2020-07-24 | 北京华宇信息技术有限公司 | Method and device for detecting blank page and fuzzy page of electronic file |
| CN111899132A (en) * | 2020-08-28 | 2020-11-06 | 四川省广安市人民检察院 | Method for automatically identifying case not setting up case within specified period |
| CN112102117A (en) * | 2020-08-17 | 2020-12-18 | 武汉市润普网络科技有限公司 | Electronic file quality inspection and supervision system |
| CN112612893A (en) * | 2020-12-29 | 2021-04-06 | 广西安怡臣信息技术有限公司 | Electronic file case generation system |
| CN113204610A (en) * | 2021-05-06 | 2021-08-03 | 广东博维创远科技有限公司 | Automatic cataloguing method based on criminal case electronic file and computer readable storage device |
| CN113220430A (en) * | 2021-04-28 | 2021-08-06 | 上海交大慧谷通用技术有限公司 | Method and system for uploading and identifying electronic file materials in parallel |
| CN113222417A (en) * | 2021-05-17 | 2021-08-06 | 广西安怡臣信息技术有限公司 | Electronic file data factory full-process intelligent application management system |
| CN116433494A (en) * | 2023-04-19 | 2023-07-14 | 南通大学 | File scanning image automatic correction and trimming method based on deep learning |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105678612A (en) * | 2015-12-30 | 2016-06-15 | 远光软件股份有限公司 | Mobile terminal original certificate electronic intelligent filling system and method |
| CN106250830A (en) * | 2016-07-22 | 2016-12-21 | 浙江大学 | Digital book structured analysis processing method |
| CN108182252A (en) * | 2017-12-29 | 2018-06-19 | 方正璞华软件（武汉）股份有限公司 | A kind of Personnel disputes arbitration information storage system and method |
| US20180336405A1 (en) * | 2017-05-17 | 2018-11-22 | Tab2Ex, Llc | Method of digitizing and extracting meaning from graphic objects |
| CN109409421A (en) * | 2018-10-09 | 2019-03-01 | 杭州诚道科技股份有限公司 | Motor vehicle, driver's archival image recognition methods based on convolutional neural networks |
-
2019
- 2019-05-27 CN CN201910448250.4A patent/CN110209632A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105678612A (en) * | 2015-12-30 | 2016-06-15 | 远光软件股份有限公司 | Mobile terminal original certificate electronic intelligent filling system and method |
| CN106250830A (en) * | 2016-07-22 | 2016-12-21 | 浙江大学 | Digital book structured analysis processing method |
| US20180336405A1 (en) * | 2017-05-17 | 2018-11-22 | Tab2Ex, Llc | Method of digitizing and extracting meaning from graphic objects |
| CN108182252A (en) * | 2017-12-29 | 2018-06-19 | 方正璞华软件（武汉）股份有限公司 | A kind of Personnel disputes arbitration information storage system and method |
| CN109409421A (en) * | 2018-10-09 | 2019-03-01 | 杭州诚道科技股份有限公司 | Motor vehicle, driver's archival image recognition methods based on convolutional neural networks |
Non-Patent Citations (1)
| Title |
|---|
| 李世普: "《生物医用材料导论》", 31 August 2000, 武汉工业大学出版社 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110597818B (en) * | 2019-09-19 | 2023-11-14 | 腾讯科技（深圳）有限公司 | File inquiry method, device, equipment and storage medium based on block chain |
| CN110597818A (en) * | 2019-09-19 | 2019-12-20 | 腾讯科技（深圳）有限公司 | Block chain based volume query method, device, equipment and storage medium |
| CN110675121A (en) * | 2019-09-23 | 2020-01-10 | 珠海市新德汇信息技术有限公司 | Method for collecting picture type file material |
| CN110675289A (en) * | 2019-09-29 | 2020-01-10 | 四川大学 | Method for compiling electronic file catalogue with case criminal review |
| CN110675289B (en) * | 2019-09-29 | 2023-05-05 | 四川大学 | Method for cataloging electronic file along with criminal investigation |
| CN110728240A (en) * | 2019-10-14 | 2020-01-24 | 北京华宇信息技术有限公司 | Method and device for automatically identifying title of electronic file |
| CN111445433A (en) * | 2019-10-14 | 2020-07-24 | 北京华宇信息技术有限公司 | Method and device for detecting blank page and fuzzy page of electronic file |
| CN112102117A (en) * | 2020-08-17 | 2020-12-18 | 武汉市润普网络科技有限公司 | Electronic file quality inspection and supervision system |
| CN111899132A (en) * | 2020-08-28 | 2020-11-06 | 四川省广安市人民检察院 | Method for automatically identifying case not setting up case within specified period |
| CN111899132B (en) * | 2020-08-28 | 2024-01-09 | 四川省广安市人民检察院 | Method for automatically identifying case not found in specified period |
| CN112612893A (en) * | 2020-12-29 | 2021-04-06 | 广西安怡臣信息技术有限公司 | Electronic file case generation system |
| CN113220430A (en) * | 2021-04-28 | 2021-08-06 | 上海交大慧谷通用技术有限公司 | Method and system for uploading and identifying electronic file materials in parallel |
| CN113204610A (en) * | 2021-05-06 | 2021-08-03 | 广东博维创远科技有限公司 | Automatic cataloguing method based on criminal case electronic file and computer readable storage device |
| CN113222417A (en) * | 2021-05-17 | 2021-08-06 | 广西安怡臣信息技术有限公司 | Electronic file data factory full-process intelligent application management system |
| CN116433494A (en) * | 2023-04-19 | 2023-07-14 | 南通大学 | File scanning image automatic correction and trimming method based on deep learning |
| CN116433494B (en) * | 2023-04-19 | 2024-02-02 | 南通大学 | File scanning image automatic correction and trimming method based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110209632A (en) | A kind of electronics folder with case production, turn shelves system | |
| Ke et al. | A review of methods for automatic individual tree-crown detection and delineation from passive remote sensing | |
| Yousif et al. | Animal Scanner: Software for classifying humans, animals, and empty frames in camera trap images | |
| KR102220174B1 (en) | Learning-data enhancement device for machine learning model and method for learning-data enhancement | |
| CN111368895B (en) | Garbage bag target detection method and detection system in wet garbage | |
| CN105844275B (en) | The localization method of line of text in text image | |
| CN110414927A (en) | A kind of bill processing automatically generates the method and device of voucher | |
| CN111488911B (en) | Image entity extraction method based on Mask R-CNN and GAN | |
| CN109460735A (en) | Document binary processing method, system, device based on figure semi-supervised learning | |
| CN109190611A (en) | Pedigree system makes are compiled in a kind of internet based on crowdsourcing | |
| CN110929746A (en) | Electronic file title positioning, extracting and classifying method based on deep neural network | |
| US11436852B2 (en) | Document information extraction for computer manipulation | |
| CN110232046A (en) | A kind of electronics folder is with case production method | |
| CN105183950B (en) | A kind of method and system for consulting engineering drawing based on mobile terminal | |
| Surana et al. | Text Extraction and Detection from Images using Machine Learning Techniques: A Research Review | |
| CN115273114A (en) | Document image optimization method and medium | |
| CN112330562B (en) | Heterogeneous remote sensing image transformation method and system | |
| CN105930391A (en) | Update method and image server of image sample database of super-resolution image system | |
| CN109190637A (en) | A kind of image characteristic extracting method | |
| CN111325207A (en) | Bill identification method and device based on preprocessing | |
| CN114066861B (en) | Coal gangue identification method based on intersection algorithm edge detection theory and visual characteristics | |
| CN110569871A (en) | saddle point identification method based on deep convolutional neural network | |
| CN104978579B (en) | The method and device that a kind of pair of image spam email is filtered | |
| Zhang et al. | Text extraction for historical Tibetan document images based on connected component analysis and corner point detection | |
| CN113361530A (en) | Image semantic accurate segmentation and optimization method using interaction means |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190906 |