CN109117938A - A kind of image scan method and system based on artificial neural network - Google Patents
A kind of image scan method and system based on artificial neural network Download PDFInfo
- Publication number
- CN109117938A CN109117938A CN201811248904.0A CN201811248904A CN109117938A CN 109117938 A CN109117938 A CN 109117938A CN 201811248904 A CN201811248904 A CN 201811248904A CN 109117938 A CN109117938 A CN 109117938A
- Authority
- CN
- China
- Prior art keywords
- neural network
- artificial
- image
- model
- network model
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/04—Scanning arrangements, i.e. arrangements for the displacement of active reading or reproducing elements relative to the original or reproducing medium, or vice versa
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Collating Specific Patterns (AREA)
Abstract
The present invention relates to artificial neural network fields, more particularly to a kind of image scan method and system based on artificial neural network, import training sample data, then the training of sample is carried out, the building of artificial nerve network model is carried out again, neural network algorithm is trained using training sample figure after the completion of artificial nerve network model building, then operation is carried out to image using neural network algorithm, after carrying out operation to image using neural network algorithm, it obtains verifying artificial nerve network model, if verification result meets default expectation quality, artificial nerve network model is verified, judge whether validation error meets default expectation quality, if verification result does not meet default expectation quality, it then jumps back to and neural network algorithm is trained again using training sample figure, until validation error meets default expectation quality, then it saves Model.The present invention can solve the problem of sweep parameter combination for being difficult to obtain linear array CCD camera optimum image.
Description
Technical field
The present invention relates to artificial neural network fields, and in particular to a kind of image scan method based on artificial neural network
And system.
Background technique
There are many factor for influencing linear array CCD camera scan image quality, determine that sweep parameter system has been difficult to set up essence
True mathematical model, therefore have to rely on the experience accumulated during many experiments for the high-quality scan image of acquisition and come
It is scanned parameter adjustment.Traditional method according to scan image prediction of quality sweep parameter is based on statistical technique
, mainly linear model.Linear model is intuitive simple, explanatory strong, but for the sweep parameter system of evolutionary series complexity
Prediction does not often prove effective, and the sweep parameter mobility of especially influence image quality factors is strong, and there are nonlinearities, traditional
Prediction technique is dealt with will be highly difficult.And selection sweep parameter is related to largely combining, in face of the conjunction of so multiple groups such as
Fruit relies only on empirical data and is difficult to obtain the sweep parameter combination of optimum image.And artificial neural network has stronger robustness
And fault-tolerance, non-linear mapping capability is suitble to solve non-linear, complication system modeling and prediction, for this purpose, we design one kind
Image scan method and system based on artificial neural network.
Summary of the invention
In view of the deficiencies of the prior art, the present invention provides a kind of image scan method based on artificial neural network and it is
System solves the problems, such as the sweep parameter combination for being difficult to obtain linear array CCD camera optimum image.
The present invention is achieved by the following technical programs:
A kind of image scan method and system based on artificial neural network, which comprises import number of training
According to；It obtains comprising file band scan image；Neural network model is verified；Preservation model, the system comprises: acquisition module,
Searching module, detection module and authentication module.
Preferably, the importing training sample data, then carry out the training of sample, then carry out artificial nerve network model
Building, neural network algorithm is trained using training sample figure after the completion of artificial nerve network model building, it is then sharp
Operation is carried out to image with neural network algorithm.
Preferably, described after carrying out operation to image using neural network algorithm, it obtains to artificial nerve network model
Verifying, judges whether the validation error meets default expectation quality, will if the verification result meets default expectation quality
The primary artificial nerve network model is determined as final artificial nerve network model, then preservation model.
Preferably, described that artificial nerve network model is verified, judge whether the validation error meets default expectation essence
Degree, if the verification result does not meet default expectation quality, then jump back to using training sample figure to neural network algorithm into
Row training, until the validation error meets default expectation quality, then preservation model.
Preferably, then the importing training sample data are obtained comprising file band scan image, are then therefrom obtained defeated
Incoming vector and output vector are removed verifying sample by input vector and output vector, are directly carried out to artificial nerve network model
Verifying.
Preferably, it is described using neural network algorithm to image carry out operation after, obtain testing artificial nerve network model
Card, judges whether the validation error meets default expectation quality, if the verification result meets default expectation quality, by institute
It states primary artificial nerve network model and is determined as final artificial nerve network model, then preservation model, if the verifying is tied
Fruit does not meet default expectation quality, then jumps back to and be trained using training sample figure to neural network algorithm again, until described
Validation error meets default expectation quality, then preservation model.
The invention has the benefit that the present invention safe and reliable can solve to be difficult to obtain the optimal figure of linear array CCD camera
Then the problem of sweep parameter combination of picture, is verified neural network model, judgement by obtaining comprising file band scan image
Whether the validation error meets default expectation quality, if the verification result meets default expectation quality, by the primary
Artificial nerve network model is determined as final artificial nerve network model, then preservation model, if the verification result is not inconsistent
Default expectation quality is closed, then jumps back to and neural network algorithm is trained again using training sample figure, until the verifying misses
Difference meets default expectation quality, then preservation model, and by obtaining module, searching module, detection module and authentication module pair
The sweep parameter combination of neural network established simpler and systematization and get optimum image.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is the flow chart of invention；
Fig. 2 is the structural schematic diagram of invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
Referring to Fig. 1-2: a kind of image scan method and system based on artificial neural network, which comprises import
Training sample data；It obtains comprising file band scan image；Neural network model is verified；Preservation model, the system comprises:
Obtain module, searching module, detection module and authentication module.
Specifically, the importing training sample data, then carry out the training of sample, then carry out artificial nerve network model
Building, neural network algorithm is trained using training sample figure after the completion of artificial nerve network model building, it is then sharp
Operation is carried out to image with neural network algorithm.
Specifically, it is described after carrying out operation to image using neural network algorithm, it obtains to artificial nerve network model
Verifying, judges whether the validation error meets default expectation quality, will if the verification result meets default expectation quality
The primary artificial nerve network model is determined as final artificial nerve network model, then preservation model.
Specifically, described verify artificial nerve network model, judge whether the validation error meets default expectation essence
Degree, if the verification result does not meet default expectation quality, then jump back to using training sample figure to neural network algorithm into
Row training, until the validation error meets default expectation quality, then preservation model.
Specifically, the importing training sample data, then obtain comprising file band scan image, then therefrom obtain defeated
Incoming vector and output vector are removed verifying sample by input vector and output vector, are directly carried out to artificial nerve network model
Verifying.
Specifically, it is described using neural network algorithm to image carry out operation after, obtain testing artificial nerve network model
Card, judges whether the validation error meets default expectation quality, if the verification result meets default expectation quality, by institute
It states primary artificial nerve network model and is determined as final artificial nerve network model, then preservation model, if the verifying is tied
Fruit does not meet default expectation quality, then jumps back to and be trained using training sample figure to neural network algorithm again, until described
Validation error meets default expectation quality, then preservation model.
The present invention safe and reliable can solve the sweep parameter combination for being difficult to obtain linear array CCD camera optimum image
Then problem verifies neural network model by obtaining comprising file band scan image, judges whether the validation error accords with
Default expectation quality is closed, it is if the verification result meets default expectation quality, the primary artificial nerve network model is true
It is set to final artificial nerve network model, then preservation model, if the verification result does not meet default expectation quality, then
It jumps back to and neural network algorithm is trained using training sample figure, until the validation error meets default expectation quality,
Then preservation model, and it is simpler to the foundation of neural network by obtaining module, searching module, detection module and authentication module
The sweep parameter combination of optimum image is got with systematization.
The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although with reference to the foregoing embodiments
Invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each implementation
Technical solution documented by example is modified or equivalent replacement of some of the technical features；And these modification or
Replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution.
Claims (6)
1. a kind of image scan method and system based on artificial neural network, it is characterised in that: the described method includes: importing instruction
Practice sample data；It obtains comprising file band scan image；Neural network model is verified；Preservation model, the system comprises: it obtains
Modulus block, searching module, detection module and authentication module.
2. a kind of image scan method and system based on artificial neural network according to claim 1, it is characterised in that:
The importing training sample data, then carry out the training of sample, then carry out the building of artificial nerve network model, artificial neuron
Neural network algorithm is trained using training sample figure after the completion of network model building, then utilizes neural network algorithm pair
Image carries out operation.
3. a kind of image scan method and system based on artificial neural network according to claim 1, it is characterised in that:
It is described after carrying out operation to image using neural network algorithm, obtain verifying artificial nerve network model, test described in judgement
Whether card error meets default expectation quality, if the verification result meets default expectation quality, by the primary artificial mind
It is determined as final artificial nerve network model through network model, then preservation model.
4. a kind of image scan method and system based on artificial neural network according to claim 1, it is characterised in that:
It is described that artificial nerve network model is verified, judge whether the validation error meets default expectation quality, if the verifying is tied
Fruit does not meet default expectation quality, then jumps back to and be trained using training sample figure to neural network algorithm again, until described
Validation error meets default expectation quality, then preservation model.
5. a kind of image scan method and system based on artificial neural network according to claim 1, it is characterised in that:
The importing training sample data, then obtain comprising file band scan image, then therefrom obtain input vector and export to
Amount is removed verifying sample by input vector and output vector, is directly verified to artificial nerve network model.
6. a kind of image scan method and system based on artificial neural network according to claim 1, it is characterised in that:
It is described using neural network algorithm to image carry out operation after, obtain to artificial nerve network model verify, judge the verifying
Whether error meets default expectation quality, if the verification result meets default expectation quality, by the primary artificial neuron
Network model is determined as final artificial nerve network model, then preservation model, if the verification result does not meet the default phase
Hope precision, then jump back to and neural network algorithm be trained again using training sample figure, until the validation error meet it is pre-
If expectation quality, then preservation model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811248904.0A CN109117938A (en) | 2018-10-25 | 2018-10-25 | A kind of image scan method and system based on artificial neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811248904.0A CN109117938A (en) | 2018-10-25 | 2018-10-25 | A kind of image scan method and system based on artificial neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109117938A true CN109117938A (en) | 2019-01-01 |
Family
ID=64855255
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811248904.0A Pending CN109117938A (en) | 2018-10-25 | 2018-10-25 | A kind of image scan method and system based on artificial neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109117938A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112883670A (en) * | 2021-03-17 | 2021-06-01 | 电子科技大学 | Inductance automatic design comprehensive model and method based on artificial neural network |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6301381B1 (en) * | 1995-03-07 | 2001-10-09 | Matsushita Electric Idustrial Co., Ltd. | Neurofilter, and method of training same to operate on image data such as to discriminate between text and picture regions of an image which is expressed by image data |
| CN101807245A (en) * | 2010-03-02 | 2010-08-18 | 天津大学 | Artificial neural network-based multi-source gait feature extraction and identification method |
| CN105354611A (en) * | 2015-10-08 | 2016-02-24 | 程涛 | Artificial neural network based best quality image scanning method and system |
| CN106096531A (en) * | 2016-05-31 | 2016-11-09 | 安徽省云力信息技术有限公司 | A kind of traffic image polymorphic type vehicle checking method based on degree of depth study |
| CN106373397A (en) * | 2016-09-28 | 2017-02-01 | 哈尔滨工业大学 | Fuzzy neural network-based remote sensing image road traffic situation analysis method |
-
2018
- 2018-10-25 CN CN201811248904.0A patent/CN109117938A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6301381B1 (en) * | 1995-03-07 | 2001-10-09 | Matsushita Electric Idustrial Co., Ltd. | Neurofilter, and method of training same to operate on image data such as to discriminate between text and picture regions of an image which is expressed by image data |
| CN101807245A (en) * | 2010-03-02 | 2010-08-18 | 天津大学 | Artificial neural network-based multi-source gait feature extraction and identification method |
| CN105354611A (en) * | 2015-10-08 | 2016-02-24 | 程涛 | Artificial neural network based best quality image scanning method and system |
| CN106096531A (en) * | 2016-05-31 | 2016-11-09 | 安徽省云力信息技术有限公司 | A kind of traffic image polymorphic type vehicle checking method based on degree of depth study |
| CN106373397A (en) * | 2016-09-28 | 2017-02-01 | 哈尔滨工业大学 | Fuzzy neural network-based remote sensing image road traffic situation analysis method |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112883670A (en) * | 2021-03-17 | 2021-06-01 | 电子科技大学 | Inductance automatic design comprehensive model and method based on artificial neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105354611B (en) | A kind of best quality image scan method and system based on artificial neural network | |
| CN105100789B (en) | A kind of method for evaluating video quality | |
| CN100559881C (en) | A kind of method for evaluating video quality based on artificial neural net | |
| CN106296669B (en) | A kind of image quality evaluating method and device | |
| CN105917353A (en) | Feature extraction and matching and template update for biometric authentication | |
| CN110781976B (en) | Extension method of training image, training method and related device | |
| CN110246137A (en) | A kind of imaging method, device and storage medium | |
| US11436774B2 (en) | Pattern mapping | |
| EP3115932A1 (en) | Image reconstruction | |
| CN115220061B (en) | Orthogonal normalization-based deep learning polarization ghost imaging method and system | |
| Quiring et al. | Fragile sensor fingerprint camera identification | |
| CN109117938A (en) | A kind of image scan method and system based on artificial neural network | |
| AU2017208235A1 (en) | Relative position encoding based networks for action recognition | |
| CN111141653A (en) | Tunnel leakage rate prediction method based on neural network | |
| CN109815959A (en) | A kind of Yield Forecast of Winter Wheat method and device | |
| CN114299573A (en) | Video processing method and device, electronic equipment and storage medium | |
| CN110490796A (en) | A kind of human face super-resolution processing method and system of the fusion of low-and high-frequency ingredient | |
| CN110210548A (en) | A kind of picture dynamic self-adapting compression method based on intensified learning | |
| CN111667409B (en) | Super-resolution algorithm-based insulator image resolution enhancement method | |
| CN112183224A (en) | Model training method for image recognition, image recognition method and device | |
| CN116895100A (en) | Knowledge distillation depth counterfeiting detection method and system based on space-frequency feature fusion | |
| CN110517234A (en) | Feature bone method for detecting abnormality and device | |
| US10395129B2 (en) | Dynamic registration seed | |
| US20150296136A1 (en) | Method and a procedure to replicate visual acuity deficiencies for use in accessibility verification | |
| US10691919B1 (en) | Dynamic registration using multiple match enrollment |
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 |
Application publication date: 20190101 |
|
| RJ01 | Rejection of invention patent application after publication |