CN108492337A - A kind of gray level image simplification method, unit and storage medium based on PCA - Google Patents
A kind of gray level image simplification method, unit and storage medium based on PCA Download PDFInfo
- Publication number
- CN108492337A CN108492337A CN201810123063.4A CN201810123063A CN108492337A CN 108492337 A CN108492337 A CN 108492337A CN 201810123063 A CN201810123063 A CN 201810123063A CN 108492337 A CN108492337 A CN 108492337A
- Authority
- CN
- China
- Prior art keywords
- gray level
- matrix
- level image
- compression
- obtains
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- 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
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T9/00—Image coding
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Compression Of Band Width Or Redundancy In Fax (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of, and the gray level image based on PCA simplifies method, the method includes the steps:Extract the gray level image matrix of gray level image;It compresses the gray level image matrix and obtains first compression gray level image matrix;It compresses the first compression gray level image matrix and obtains second-compressed gray level image matrix;Restore the gray level image after the second-compressed gray level image matrix is simplified.The present invention also provides a kind of active interactive device, non-transient readable storage medium storing program for executing and devices, for realizing the method.The present invention can reduce the redundancy of gray level image, reduce gray level image the space occupied in the electronic device.
Description
Technical field
The present invention relates to image processing fields, simplify method more particularly, to a kind of gray level image based on PCA, set
Standby, device and storage medium.
Background technology
In the research and application in many fields, generally require to reflecting that multiple variables of things are largely observed,
It collects mass data and finds rule to carry out analysis.Multivariable large sample undoubtedly can provide abundant letter for research and application
Breath, but the workload of data acquisition is also increased to a certain extent.If analyzed respectively each index, analysis is often
It is isolated, rather than comprehensive.Many information can be lost by blindly reducing index, easy to produce the conclusion of mistake.Therefore it needs
A rational method is found, the index analyzed is needed simultaneously in reduction, the loss that former index includes information is reduced to the greatest extent, to reach
To the purpose analyzed collected data comprehensively.Therefore, a kind of redundancy that can reduce image is found, is realized to figure
The compression of picture, while the method that the minimum error of image may be implemented by retrieving algorithm just become the problem of industry is paid close attention to.
Invention content
It solves the above problems in order to overcome the problems referred above or at least partly, the present invention provides a kind of ashes based on PCA
Spend image simplification method, unit and storage medium.
On the one hand, the present invention provides a kind of, and the gray level image based on PCA simplifies method, and specific steps include:Extraction ash
Spend the gray level image matrix of image;It compresses the gray level image matrix and obtains first compression gray level image matrix;Compress described one
Second compression gray level image matrix obtains second-compressed gray level image matrix;It restores the second-compressed gray level image matrix and obtains letter
Gray level image after change.
On the other hand, the present invention provides a kind of active interactive device and a kind of non-transient readable storage medium storing program for executing.Described one
Planting active interactive device includes:At least one processor;And at least one processor being connect with the processor communication,
In:The memory is stored with the program instruction that can be executed by the processor, and the processor calls described program to instruct energy
It enough executes a kind of gray level image based on PCA and simplifies method.A kind of non-transient readable storage medium storing program for executing storage program refers to
It enables, simplifies method for executing a kind of gray level image based on PCA.The present invention also provides a kind of devices, for realizing
The method.
The present invention provides a kind of, and the gray level image based on PCA simplifies method, unit and storage medium, by adopting
Gray level image is compressed with PCA methods, it is possible to reduce the redundancy of gray level image reduces gray level image in electronic equipment
Middle the space occupied.
Description of the drawings
Fig. 1 is the overall flow figure that the gray level image based on PCA simplifies method in the embodiment of the present invention;
Fig. 2 is to compress gray level image matrix in the embodiment of the present invention to obtain first compression gray level image matrix specific steps stream
Cheng Tu;
Fig. 3 is to compress first compression gray level image matrix in the embodiment of the present invention to obtain second-compressed gray level image matrix tool
Body flow chart of steps;
Fig. 4 is the gray level image effect diagram after simplification in the embodiment of the present invention;
Fig. 5 is gray level image matrix and original image matrix schematic diagram in the embodiment of the present invention;
Fig. 6 is the hardware device operating diagram of the embodiment of the present invention;
Fig. 7 is that the gray level image based on PCA simplifies schematic device in the embodiment of the present invention.
Specific implementation mode
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention
Formula is further described, and the particular technique details hereinafter mentioned only makes reader be better understood from technical solution, not generation
Present invention is limited only by following technical details for table.
The embodiment provides a kind of, and the gray level image based on PCA simplifies method, unit and storage Jie
Matter.It is the overall flow figure that the gray level image based on PCA simplifies method in the embodiment of the present invention, the method referring to Fig. 1, Fig. 1
It is realized by hardware device, specific steps include:
S101:The gray level image matrix of gray level image is extracted, the gray level image includes:Appointing in three primary colours gray level image
It anticipates a kind of or arbitrary several combination, the element type of the gray level image matrix includes:Floating-point type double precision.
S102:It compresses the gray level image matrix and obtains first compression gray level image matrix.
S103:It compresses the first compression gray level image matrix and obtains second-compressed gray level image matrix.
S104:Restore the gray level image after the second-compressed gray level image matrix is simplified.
It is to compress gray level image matrix in the embodiment of the present invention to obtain first compression gray level image matrix tool referring to Fig. 2, Fig. 2
Body flow chart of steps, including:
S201:Calculate mean value of the gray level image matrix per column element.
S202:Each element of the gray level image matrix subtracts the element mean value of column, obtains equal value difference matrix.
S203:Principal component analysis is carried out to the equal value difference matrix, obtains the characteristic value and feature of the equal value difference matrix
Vector.
S204:The contribution rate individual amount threshold value for setting principal component analysis, according to the contribution rate individual amount threshold value, choosing
The feature vector composition characteristic vector matrix for taking the equal value difference matrix, specifically includes:The contribution rate individual amount threshold value is
100, according to the descending sequence of contribution rate, choose in equal value difference matrix before contribution rate 100 100 feature vector composition characteristics
Vector matrix.
S205:The equal value difference matrix right side multiplies described eigenvector matrix, obtains first compression gray level image matrix.
It is to compress first compression gray level image matrix in the embodiment of the present invention to obtain second-compressed gray-scale map referring to Fig. 3, Fig. 3
As matrix specific steps flow chart, including:
S301:First compression gray level image matrix described in transposition obtains transposed matrix, calculates the every column element of transposed matrix
Mean value.
S302:Each element of the transposed matrix subtracts the element mean value of column, obtains the equal value difference matrix of transposition.
S303:Principal component analysis is carried out to the equal value difference matrix of the transposition, obtains the feature of the equal value difference matrix of the transposition
Value and feature vector.
S304:The contribution rate individual amount critical value of principal component analysis is set, it is critical according to the contribution rate individual amount
Value is chosen the feature vector composition transposition eigenvectors matrix of the equal value difference matrix of the transposition, is specifically included:The contribution rate
Body quantity critical value is 60, according to the descending sequence of contribution rate, chooses in the equal value difference matrix of transposition 60 of before contribution rate 60
Feature vector forms transposition eigenvectors matrix.
S305:The transposition equal value difference matrix right side multiplies the transposition eigenvectors matrix, obtains second-compressed gray level image
Matrix.
It is described to restore the gray level image after the second-compressed gray level image matrix is simplified, it specifically includes:Described two
The second compression gray level image matrix right side multiplies the transposed matrix of the transposition eigenvectors matrix, obtains inverse matrix;The inverse matrix is right
The transposed matrix for multiplying described eigenvector matrix, the gray level image after being simplified.
It is the gray level image effect diagram in the embodiment of the present invention after simplification referring to Fig. 4, Fig. 4, including:Original graph
Simplify ash as 401, red primary gray level image 402, red primary simplify gray level image 403, green primary gray level image 404, green primary
It spends image 405, blue primary gray level image 406, blue primary and simplifies gray level image 407.Original image 401 is decomposed into red primary gray scale
Image 402, green primary gray level image 404 and blue primary gray level image 406.Red primary gray level image 402 is after overcompression restores
It obtains red primary and simplifies gray level image 403.Green primary gray level image 404 obtains green primary after overcompression restores and simplifies gray-scale map
As 405.Blue primary gray level image 406 obtains blue primary after overcompression restores and simplifies gray level image 407.It can be seen that simple
Element after change in image substantially reduces, and compression ratio reaches 829.47.
It is gray level image matrix and original image matrix schematic diagram in the embodiment of the present invention referring to Fig. 5, Fig. 5, including:Red base
Color shade image array 501, original image matrix 502, element minimum value 503 and element maximum value 504.Wherein, 520x520 tables
Show and share 270400 elements, 520x520x3 indicates 811200 elements, wherein 3 indicate that three primary colours, uint8 indicate element
Data type, element minimum value 503 indicates in 270400 elements in red primary gray level image matrix 501, minimum
Element value is 3.In 270400 elements of the expression of element maximum value 504 in red primary gray level image matrix 501, maximum member
Plain value is 255.
It is the hardware device operating diagram of the embodiment of the present invention referring to Fig. 6, Fig. 6, the hardware device specifically includes:One
Gray level image of the kind based on PCA simplifies equipment 601, processor 602 and storage medium 603.
Gray level image based on PCA simplifies equipment 601:A kind of gray level image based on PCA simplifies equipment 601 and realizes
A kind of gray level image based on PCA simplifies method.
Processor 602:The processor 602 loads and executes the instruction in the storage medium 603 and data for real
A kind of existing gray level image based on PCA simplifies method.
Storage medium 603:603 store instruction of the storage medium and data;The storage medium 603 is for realizing described
A kind of gray level image based on PCA simplify method.
It is the gray level image simplification schematic device based on PCA in the embodiment of the present invention referring to Fig. 7, Fig. 7, including:Gray scale
Image array extraction module 701, first compression module 702, second-compressed module 703 and compression recovery module 704;The gray scale
Image array extraction module 701 extracts the gray level image matrix of gray level image, and by gray level image Transfer-matrix to described primary
Compression module 702;The first compression module 702 compresses gray level image matrix and obtains first compression gray level image matrix, and will
The first compression gray level image Transfer-matrix gives the second-compressed module 703;The second-compressed module 703 is compressed described
First compression gray level image matrix obtains second-compressed gray level image matrix, and by the second-compressed gray level image Transfer-matrix
To the compression recovery module 704, the compression recovery module 704 restores the second-compressed gray level image matrix and is simplified
Gray level image afterwards
By executing the embodiment of the present invention, all technical characteristics in the claims in the present invention are obtained for detailed explain
It states.
It is different from the prior art, the embodiment provides a kind of, and the gray level image based on PCA simplifies method, sets
Standby, device and storage medium compress gray level image by using PCA methods, it is possible to reduce the redundancy of gray level image is believed
Breath reduces gray level image the space occupied in the electronic device.
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent replacement, improvement and so on should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of gray level image based on PCA simplifies method, which is characterized in that including:
Step 1:Extract the gray level image matrix of gray level image;
Step 2:It compresses the gray level image matrix and obtains first compression gray level image matrix;
Step 3:It compresses the first compression gray level image matrix and obtains second-compressed gray level image matrix;
Step 4:Restore the gray level image after the second-compressed gray level image matrix is simplified.
2. the method as described in claim 1, which is characterized in that the gray level image includes:Appointing in three primary colours gray level image
It anticipates a kind of or arbitrary several combination.
3. method as claimed in claim 1 or 2, which is characterized in that the compression gray level image matrix is once pressed
Contracting gray level image matrix, including:
Step 1:Calculate mean value of the gray level image matrix per column element;
Step 2:Each element of the gray level image matrix subtracts the element mean value of column, obtains equal value difference matrix;
Step 3:Principal component analysis is carried out to the equal value difference matrix, obtain the equal value difference matrix characteristic value and feature to
Amount;
Step 4:, according to contribution rate individual amount threshold value, choose the feature vector composition characteristic moment of a vector of the equal value difference matrix
Battle array;
Step 5:The equal value difference matrix right side is multiplied into described eigenvector matrix, obtains first compression gray level image matrix.
4. method as claimed in claim 3, which is characterized in that the contribution rate individual amount threshold value is 100;It is described according to tribute
Rate individual amount threshold value is offered, the feature vector composition characteristic vector matrix of the equal value difference matrix is chosen, specifically includes:According to tribute
The descending sequence of rate is offered, 100 feature vector composition characteristic vector matrixs before contribution rate are chosen in equal value difference matrix.
5. method as claimed in claim 3, which is characterized in that the compression first compression gray level image matrix obtains two
Second compression gray level image matrix, specific steps include:
Step 1:First compression gray level image matrix described in transposition obtains transposed matrix, and it is equal per column element to calculate transposed matrix
Value;
Step 2:The element mean value that each element of the transposed matrix is subtracted to column obtains the equal value difference matrix of transposition;
Step 3:To the equal value difference matrix of the transposition carry out principal component analysis, obtain the equal value difference matrix of the transposition characteristic value and
Feature vector;
Step 4:According to contribution rate individual amount critical value, the feature vector composition transposition for choosing the equal value difference matrix of the transposition is special
Levy vector matrix;
Step 5:The transposition equal value difference matrix right side is multiplied into the transposition eigenvectors matrix, obtains second-compressed gray level image square
Battle array.
6. method as claimed in claim 5, which is characterized in that the contribution rate individual amount critical value is 60;The basis
Contribution rate individual amount critical value chooses the feature vector composition transposition eigenvectors matrix of the equal value difference matrix of the transposition, tool
Body includes:, according to the descending sequence of contribution rate, 60 feature vectors compositions turn before contribution rate in the selection equal value difference matrix of transposition
Set eigenvectors matrix.
7. method as claimed in claim 5, which is characterized in that the reduction second-compressed gray level image matrix obtains letter
Gray level image after change, specifically includes:The second-compressed gray level image matrix right side multiplies turning for the transposition eigenvectors matrix
Matrix is set, inverse matrix is obtained;The inverse matrix right side multiplies the transposed matrix of described eigenvector matrix, the gray-scale map after being simplified
Picture.
8. a kind of active interactive device, which is characterized in that including:
At least one processor;And
At least one processor being connect with the processor communication, wherein:
The memory is stored with the program instruction that can be executed by the processor, and the processor calls described program to instruct energy
It is enough to execute such as any the method for claim 1 to 7.
9. a kind of non-transient readable storage medium storing program for executing, which is characterized in that the non-transient readable storage medium storing program for executing stores program instruction, institute
State program instruction for execute such as any the method for claim 1 to 7.
10. a kind of device being used for realizing any the method for claim 1 to 7, which is characterized in that including:Gray level image square
Battle array extraction module, first compression module, second-compressed module and compression recovery module;The gray level image matrix extraction module carries
The gray level image matrix of gray level image is taken, and gives gray level image Transfer-matrix to the first compression module;The first compression
Module compression gray level image matrix obtains first compression gray level image matrix, and by the first compression gray level image Transfer-matrix
To the second-compressed module;The second-compressed module compresses the first compression gray level image matrix and obtains second-compressed ash
Image array is spent, and gives the compression recovery module, the compression to restore mould the second-compressed gray level image Transfer-matrix
Block restores the gray level image after the second-compressed gray level image matrix is simplified.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810123063.4A CN108492337B (en) | 2018-02-07 | 2018-02-07 | PCA-based gray scale image simplification method, device, apparatus and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810123063.4A CN108492337B (en) | 2018-02-07 | 2018-02-07 | PCA-based gray scale image simplification method, device, apparatus and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108492337A true CN108492337A (en) | 2018-09-04 |
CN108492337B CN108492337B (en) | 2020-07-03 |
Family
ID=63344693
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810123063.4A Expired - Fee Related CN108492337B (en) | 2018-02-07 | 2018-02-07 | PCA-based gray scale image simplification method, device, apparatus and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108492337B (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103955676A (en) * | 2014-05-12 | 2014-07-30 | 苏州大学 | Human face identification method and system |
CN104616000A (en) * | 2015-02-27 | 2015-05-13 | 苏州大学 | Human face recognition method and apparatus |
CN104820696A (en) * | 2015-04-29 | 2015-08-05 | 山东大学 | Large-scale image retrieval method based on multi-label least square Hash algorithm |
-
2018
- 2018-02-07 CN CN201810123063.4A patent/CN108492337B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103955676A (en) * | 2014-05-12 | 2014-07-30 | 苏州大学 | Human face identification method and system |
CN104616000A (en) * | 2015-02-27 | 2015-05-13 | 苏州大学 | Human face recognition method and apparatus |
CN104820696A (en) * | 2015-04-29 | 2015-08-05 | 山东大学 | Large-scale image retrieval method based on multi-label least square Hash algorithm |
Also Published As
Publication number | Publication date |
---|---|
CN108492337B (en) | 2020-07-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Narayanaperumal et al. | VLSI Implementations of Compressive Image Acquisition using Block Based Compression Algorithm. | |
CA2519358C (en) | Color image compression via spectral decorrelation and elimination of spatial redundancy | |
CN105933708B (en) | A kind of method and apparatus of data compression and decompression | |
CN107743235B (en) | Image processing method, device and electronic equipment | |
Shen et al. | Joint sparse representation model for multi-channel image based on reduced geometric algebra | |
US20190325557A1 (en) | Image processing device, image processing method, and image processing program | |
CN105100814A (en) | Methods and devices for image encoding and decoding | |
Pan et al. | Enhanced fast encoding method for vector quantization by finding an optimally-ordered Walsh transform kernel | |
Chen et al. | Color image inpainting via robust pure quaternion matrix completion: Error bound and weighted loss | |
CN104322063B (en) | Method, system and the computer readable storage medium of video stream compression | |
Swathi et al. | Image compression using singular value decomposition | |
CN109286819A (en) | Combine explicit image encryption, decryption method and the device of compression | |
CN108492337A (en) | A kind of gray level image simplification method, unit and storage medium based on PCA | |
CN104683818B (en) | Method for compressing image based on biorthogonal invariant set m ultiwavelet | |
US20200013201A1 (en) | Method and system for solving inverse problems in image processing using deep dictionary learning (ddl) | |
CN107169925B (en) | The method for reconstructing of stepless zooming super-resolution image | |
CN107223259A (en) | Image storage system based on tree | |
Neagoe | An optimum 2d color space for pattern recognition. | |
CN115474048A (en) | Rapid color image compression method based on split quaternion model | |
CN107172425A (en) | Reduced graph generating method, device and terminal device | |
Moin et al. | Face recognition in JPEG compressed domain: a novel coefficient selection approach | |
CN108460409A (en) | A kind of fish identification method, equipment and storage medium based on PCA | |
CN110572652A (en) | Static image processing method and device | |
Kountchev et al. | Image color space transform with enhanced KLT | |
WO2015167519A1 (en) | Color extrapolation from a first color space to a second color space |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200703 Termination date: 20210207 |
|
CF01 | Termination of patent right due to non-payment of annual fee |