CN107766896A - A kind of spectrum dimension reduction method based on form and aspect cluster - Google Patents
A kind of spectrum dimension reduction method based on form and aspect cluster Download PDFInfo
- Publication number
- CN107766896A CN107766896A CN201711221418.5A CN201711221418A CN107766896A CN 107766896 A CN107766896 A CN 107766896A CN 201711221418 A CN201711221418 A CN 201711221418A CN 107766896 A CN107766896 A CN 107766896A
- Authority
- CN
- China
- Prior art keywords
- subregion
- sample
- color
- spectrum
- training sample
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- 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
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/58—Extraction of image or video features relating to hyperspectral data
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Spectrometry And Color Measurement (AREA)
Abstract
The present invention relates to a kind of spectrum dimension reduction method based on form and aspect cluster, it is characterised in that:Selection characteristic according to possessed by spectrum samples itself is set out, according to training sample color characteristics main form and aspect are determined according to hue circle principle, the training sample corresponding to each main form and aspect is determined using form and aspect cluster, then carries out the spectral reflectivity dimension-reduction treatment of sample to be tested.The present invention takes into full account the demand of practical application, has the dimensionality reduction and compression method for calculating the multispectral datas of feature such as simple, easily operated, characteristic vector is few, dimensionality reduction precision height;It is also more convenient with respect to for user's use using the high accuracy compression and storage of multispectral colouring information.
Description
Technical field
The invention belongs to the technical field across media color reproduction based on spectrum in color science technology, and in particular to arrive
The dimensionality reduction and compression method of multispectral data, especially a kind of spectrum dimension reduction method based on form and aspect cluster.
Background technology
During across media color reproductions, people often require that both reproducing color therewith and primitive color under circumstances
Between colour vision impression it is consistent, that is, realize " What You See Is What You Get " color-match target.So quantitative description of color
With match be all the time just color science important research point.Conventional color quantitative description is mainly based upon Ge Lasi with matching
Based on graceful law, by simulating the feature of human visual system, come quantitative description color in the form of three-dimensional color space,
And color-match is carried out based on the color space;This color-match is the color-match based on colourity, belongs to condition coupling, is chased after
The standard asked is the color-match under the conditions of certain Orientation observation, when observation condition or light source change, the color of its color
Perception changes, and color no longer matches.The mode that another kind quantitatively describes color is the spectral characteristic of object, be may be implemented in
Color-match under the conditions of any, belongs to unconditional matching, can realize the color-match mesh of " What You See Is What You Get " of real meaning
Mark, fundamentally ensure that the uniformity across media color reproduction.
Color reproduction technology based on spectrum is based on Spectral matching, all the time using the spectral characteristic rather than chromatic value of object come
Carry out the exchange of colouring information.Transmit, reproduce, influence of its colouring information also not dependent on specific environment of observation and light source.
It is that lifting substantially is brought to current color reproducing technology, is the real fundamental way realized across media color reproduction.
Because the spectral color information of acquisition is multidimensional data, it is used at it in color reproduction and color transmittance process,
The problem of the methods of data redundancy, computationally intensive, memory space is big be present.In many application scenarios(Such as arts reproduction,
Computer vision, ecommerce etc.), multidimensional spectrum is unsuitable directly to carry out spectrum-image procossing, across media color reproductions, light
Spectral domain maps.And spectrum reduction process is using the correlation between spectral reflectivity sample or between spectral band, drop is utilized
Dimension method by the process of the space dimensionality reduction of higher-dimension spectroscopic data to lower dimensional space, during reducing color treatments the needed time answer
Miscellaneous degree and processing complexity.It is to realize colouring information transmission with exchanging to establish towards the spectral reflectivity dimension reduction method of color reproduction
Important step.
The multispectral dimension reduction method suitable for color reproduction is not too many at present, can be substantially according to its dimension reduc-ing principle
It is divided into two major classes:One kind is the black penalty method of metamerism;Another kind is Multivariate Analysis.The black penalty method of metamerism is main
It is to be decomposed into basic spectrum and metamerism black light spectrum according to color spectrum information;Basic spectrum decides the colourity letter of color
Breath, metamerism black light, which is composed, decides the precision of spectrum, its main algorithm have LabPQR method of descents, LabRGB method of descents,
XYZLMS method of descents etc.;But the light conditions that this method must give, it is not good enough in aberration robustness when light change.Polynary system
The correlation that meter analytic approach is mainly based upon between the correlation between spectroscopic data or wave band is carried out using Multivariate Statistical Theory
Spectroscopic data dimensionality reduction, its method have PCA, Non-linear Principal Component method, independent PCA etc.;But this method
Only it is from the sample overall situation, does not account for the possessed characteristic of spectrum samples itself, cause spectrum dimensionality reduction precision low.It is based on
Above reason, the present invention propose a kind of sample properties using form and aspect cluster and carry out spectroscopic data dimensionality reduction to realize multispectral number
According to dimensionality reduction and compression.
The content of the invention
The present invention is in order to make up the deficiencies in the prior art, there is provided one kind, which has, calculates simple, easily operated, characteristic vector
Less, the dimensionality reduction and compression method of the multispectral data of the feature such as dimensionality reduction precision height, and the spectrum based on form and aspect cluster is built with this
Dimensionality reduction system.
What the present invention was achieved through the following technical solutions:
A kind of spectrum dimension reduction method based on form and aspect cluster, it is characterised in that comprise the following steps:
Step S1, main form and aspect are determined according to hue circle principle according to training sample color characteristics, determined using form and aspect cluster every
Training sample subset corresponding to individual main form and aspect;The number of partitions of gained is m, and now each subregion, which correspond to one, has phase
With the spectral value subset of color characteristic, and calculate hue angle and saturation degree possessed by each training sample.
Step S2, judge whether to divide low saturation color subregion:
(a)If low saturation color subregion is not divided, it is determined that hue-angle range computing corresponding to each subregion, according to hue-angle range computing
Subregion is marked ascending order, respectively H(1), H(2), H(3)…H(m), and light corresponding to each subregion difference
Spectrum subset is designated as A(1), A(2), A(3)…A(m).
(b)If dividing low saturation color subregion, the saturation degree threshold value corresponding to low saturation color subregion is first determined.When
Saturation degree corresponding to training sample then belongs to low saturation color subregion less than saturation degree threshold value, when saturation corresponding to training sample
Degree then belongs to other subregions not less than saturation degree threshold value, and low saturation color Labelling Regions are H(1), its corresponding spectral value collection
Conjunction is designated as A(1).The hue-angle range computing corresponding to other subregions is determined again, according to the ascending order of hue-angle range computing to it
It is marked, respectively H(2), H(3), H(4)…H(m), and corresponding spectrum value set is designated as A to each subregion respectively(2), A
(3), A(4)…A(m).
Step S3, Multivariate Analysis is used to the spectrum value set of each subregion, obtain the feature of each subregion to
Amount;The demand of actual production is taken into full account, selects dimensionality reduction dimension of the corresponding characteristic vector quantity as the subregion;
Step S4, the subregion where sample to be tested is judged using form and aspect cluster, using the characteristic vector of the subregion to be tested
Sample carries out spectrum dimension-reduction treatment;Spectral Reconstruction, the sample to be tested after being reconstructed are carried out finally by subregion where searching
Spectral reflectivity.
In a kind of spectrum dimension reduction method based on form and aspect cluster provided by the present invention, can also have the feature that:
Wherein, step S1 comprises the following steps:
Step S1-1, main form and aspect are determined according to hue circle principle according to training sample color characteristics, is clustered and determined using form and aspect
Training sample corresponding to each main form and aspect;
Step S1-2, the number of partitions of gained is m, and now each subregion correspond to a spectral value with same color feature
Set;
Step S1-3, calculate hue angle and saturation degree possessed by training sample;Specific method has two methods:
Method 1:Under fixed environment of observation, the CIE XYZ values corresponding to training sample spectral reflectivity are calculated, i.e.,:
Wherein,kTo adjust factor,For spectral reflectivity corresponding to training sample,For the relative spectral power distribution of spectrum
Curve,It is CIE color matching functions.
The CIE XYZ values of training sample are transformed into by CIE Lab values by color space conversion function, hue angle h and full
With degree C, i.e.,:
Wherein,For the CIE XYZ values of training sample,It is irradiated to for CIE standard illuminants completely unrestrained
The tristimulus values of reflector surface;
Method 2:Spectral value collection corresponding to training sample set is carried out to use Multivariate Analysis(Such as PCA, solely
Vertical PCA, non-negative PCA etc.), obtain characteristic vector;According to the variance contribution ratio of characteristic vector(It is i.e. square
Poor contribution rate represents the ratio of the information content contained under this feature vector)Mode from big to small arranged, be designated as pc
(1), pc(2), pc(3)…pc(n).
Wherein,Be itsiIndividual characteristic vector,It is characteristic vectoriCorresponding coefficient, n are spectrum sample intervals.
Take first three characteristic vector pc(1), pc(2), pc(3), the hue angle h and saturation degree C of training sample are calculated, i.e.,:
In holding constant across the media method of color gamut mapping of color of form and aspect provided by the present invention, can also have the feature that:Its
In, step S4 comprises the following steps:
Step S4-1, select the method for obtaining hue angle and saturation degree with calculating training sample consistent, calculate sample to be tested
Possessed hue angle and saturation degree;
Step S4-2, judge the subregion where sample to be tested, the subregion where sample to be tested is judged using form and aspect cluster, and
Subregion where treating test sample is marked;If sample to be tested just exists in the line of demarcation of two subregions, sample to be tested
Two subregions all carry out dimensionality reduction, calculate the color between the numerical value after dimensionality reduction in the case of converting various light sources and sample to be tested
Difference, the subregion where selecting multiple light courcess value of chromatism sum minimum mark place subregion as the subregion where sample to be tested.
Step S4-3, treat test sample using the characteristic vector of the subregion and carry out spectrum dimensionality reduction;Finally by lookup institute
Spectral Reconstruction, the sample spectral reflectivity to be tested after being reconstructed are carried out in subregion.
The beneficial effects of the invention are as follows:
The present invention takes into full account the demand of practical application, selects the characteristic according to possessed by spectrum samples itself to set out, according to instruction
Practice sample of color characteristic and determine main form and aspect according to hue circle principle, determined using form and aspect cluster corresponding to each main form and aspect
Training sample, there is the dimensionality reduction for calculating the multispectral datas of feature such as simple, easily operated, characteristic vector is few, dimensionality reduction precision height
With compression method;It is also more convenient with respect to for user's use using the high accuracy compression and storage of multispectral colouring information.
Brief description of the drawings
Fig. 1 is a kind of spectrum dimension reduction method flow chart based on form and aspect cluster of the present invention.
Embodiment
In order that the technical means, the inventive features, the objects and the advantages of the present invention are easy to understand, it is real below
Example combination accompanying drawing is applied to be specifically addressed the present invention based on the spectrum dimension reduction method that form and aspect cluster.
Fig. 1 is a kind of flow chart of the spectrum dimension reduction method based on form and aspect cluster of the present invention.
As shown in figure 1, Munsell colour system, Oswald that moral color system, Chinese color may be selected in training sample set
System etc..In the present embodiment, training sample set is treated by taking Munsell colour system as an example used in spectrum dimension reduction method
Test sample carries out spectrum dimensionality reduction, to realize the dimensionality reduction of multispectral data and compression.A kind of spectrum dimensionality reduction based on form and aspect cluster
Method comprises the following steps that:
Step S1, main form and aspect are determined according to hue circle principle according to the color characteristics of Munsell colour system, it is respectively red
(R), Huang Hong(YR), it is yellow(Y), it is greenish-yellow(GY), it is green(G), it is bluish-green(BG), it is blue(B), purplish blue(PB), it is purple(P)And purple(RP);Root
According to the requirement of actual demand, low saturation color subregion can be divided;Then each main form and aspect institute is determined using form and aspect cluster
Corresponding training sample;The number of partitions of gained is m, and now each subregion correspond to a spectrum with same color feature
Value set, and calculate hue angle possessed by training sample and saturation degree.Concrete operation step is as follows:
Step S1-1, main form and aspect are determined according to hue circle principle according to the color characteristics of Munsell colour system, it is respectively red
(R), Huang Hong(YR), it is yellow(Y), it is greenish-yellow(GY), it is green(G), it is bluish-green(BG), it is blue(B), purplish blue(PB), it is purple(P)And purple(RP);Root
According to the requirement of actual demand, low saturation color subregion can be divided;Then each main form and aspect institute is determined using form and aspect cluster
Corresponding training sample;
Step S1-2, the number of partitions of gained is m, and now each subregion correspond to a spectral value with same color feature
Set;
Step S1-3, calculate hue angle and saturation degree possessed by training sample;Specific method has two methods, selects wherein one
Kind method is calculated:
Method 1:Under fixed environment of observation, the CIE XYZ values corresponding to training sample spectral reflectivity are calculated, i.e.,:
Wherein,kTo adjust factor,For spectral reflectivity corresponding to training sample,For the relative spectral power distribution of spectrum
Curve,It is CIE color matching functions.
The CIE XYZ values of training sample are transformed into by CIE Lab values by color space conversion function, hue angle h and full
With degree C, i.e.,:
Wherein,For the CIE XYZ values of training sample,It is irradiated to for CIE standard illuminants completely unrestrained
The tristimulus values of reflector surface;
Method 2:Spectral value collection corresponding to training sample set is carried out to use Multivariate Analysis(Such as PCA, solely
Vertical PCA, non-negative PCA etc.), obtain characteristic vector;According to the variance contribution ratio of characteristic vector(It is i.e. square
Poor contribution rate represents the ratio of the information content contained under this feature vector)Mode from big to small arranged, be designated as pc
(1), pc(2), pc(3)…pc(n).
Wherein,Be itsiIndividual characteristic vector,It is characteristic vectoriCorresponding coefficient, n are spectrum sample intervals.
Take first three characteristic vector pc(1), pc(2), pc(3), the hue angle h and saturation degree C of training sample are calculated, i.e.,:
Step S2, judge whether to divide low saturation color subregion:
(a)If low saturation color subregion is not divided, it is determined that hue-angle range computing corresponding to each subregion, according to hue-angle range computing
Subregion is marked ascending order, respectively H(1), H(2), H(3)…H(m), and light corresponding to each subregion difference
Spectrum set is designated as A(1), A(2), A(3)…A(m).
(b)If dividing low saturation color subregion, the saturation degree threshold value corresponding to low saturation color subregion is first determined.When
Saturation degree corresponding to training sample then belongs to low saturation color subregion less than saturation degree threshold value, when saturation corresponding to training sample
Degree then belongs to other subregions not less than saturation degree threshold value, and low saturation color Labelling Regions are H(1), its corresponding spectral value collection
Conjunction is designated as A(1).The hue-angle range computing corresponding to other subregions is determined again, according to the ascending order of hue-angle range computing to it
It is marked, respectively H(2), H(3), H(4)…H(m), and corresponding spectrum value set is designated as A to each subregion respectively(2), A
(3), A(4)…A(m).
Step S3, Multivariate Analysis is used to the spectrum value set of each subregion, obtain the feature of each subregion to
Amount;The demand of actual production is taken into full account, selects dimensionality reduction dimension of the corresponding characteristic vector quantity as the subregion;
Step S4, the subregion where sample to be tested is judged using form and aspect cluster, using the characteristic vector of the subregion to be tested
Sample carries out spectrum dimension-reduction treatment;Spectral Reconstruction, the sample to be tested after being reconstructed are carried out finally by subregion where searching
Spectral reflectivity.Concrete operations are as follows:
Step S4-1, select the method for obtaining hue angle and saturation degree with calculating training sample consistent, calculate sample to be tested
Possessed hue angle and saturation degree;
Step S4-2, judge the subregion where sample to be tested, the subregion where sample to be tested is judged using form and aspect cluster;If
Sample to be tested just all carries out dimensionality reduction in two subregions, calculated multispectral in the line of demarcation of two subregions, sample to be tested
In the case of aberration between numerical value after dimensionality reduction and sample to be tested, the subregion where selecting multiple light courcess value of chromatism sum minimum
As the subregion where sample to be tested.
Step S4-3, treat test sample using the characteristic vector of the subregion and carry out spectrum dimensionality reduction;Finally by lookup institute
Spectral Reconstruction, the sample spectral reflectivity to be tested after being reconstructed are carried out in subregion.
The present invention takes into full account the demand of practical application, selects the characteristic according to possessed by spectrum samples itself to set out, root
Main form and aspect are determined according to hue circle principle according to training sample color characteristics, determine that each main form and aspect institute is right using form and aspect cluster
The training sample answered, have and calculate the multispectral datas of feature such as simple, easily operated, characteristic vector is few, dimensionality reduction precision height
Dimensionality reduction and compression method;It is also more square with respect to for user's use using the high accuracy compression and storage of multispectral colouring information
Just.
The effect of embodiment and effect
A kind of spectrum dimension reduction method clustered based on form and aspect provided according to the present embodiment, is selected according to spectrum samples itself institute
The characteristic having is set out, and is determined main form and aspect according to hue circle principle according to training sample color characteristics, is clustered using form and aspect true
Training sample corresponding to fixed each main form and aspect, then carries out the spectral reflectivity dimension-reduction treatment of sample to be tested.
A kind of spectrum dimension reduction method clustered based on form and aspect provided according to the present embodiment, takes into full account practical application
Demand, there is the dimensionality reduction and pressure for calculating the multispectral datas of feature such as simple, easily operated, characteristic vector is few, dimensionality reduction precision height
Compression method;It is also more convenient with respect to for user's use using the high accuracy compression and storage of multispectral colouring information.
Above-mentioned embodiment is the preferred case of the present invention, is not intended to limit protection scope of the present invention.
Claims (3)
1. a kind of spectrum dimension reduction method based on form and aspect cluster, it is characterised in that comprise the following steps:
Step S1, main form and aspect are determined according to hue circle principle according to training sample color characteristics, determined using form and aspect cluster every
Training sample subset corresponding to individual main form and aspect;The number of partitions of gained is m, and now each subregion, which correspond to one, has phase
With the spectral value subset of color characteristic, and calculate hue angle and saturation degree possessed by each training sample.
Step S2, judge whether to divide low saturation color subregion:
(a)If low saturation color subregion is not divided, it is determined that hue-angle range computing corresponding to each subregion, according to hue-angle range computing
Subregion is marked ascending order, respectively H(1), H(2), H(3)…H(m), and light corresponding to each subregion difference
Spectrum subset is designated as A(1), A(2), A(3)…A(m).
(b)If dividing low saturation color subregion, the saturation degree threshold value corresponding to low saturation color subregion is first determined.Work as training
Saturation degree corresponding to sample then belongs to low saturation color subregion less than saturation degree threshold value, when saturation degree corresponding to training sample not
Then belong to other subregions less than saturation degree threshold value, low saturation color Labelling Regions are H(1), its corresponding spectrum value set note
For A(1).The hue-angle range computing corresponding to other subregions is determined again, and it is carried out according to the ascending order of hue-angle range computing
Mark, respectively H(2), H(3), H(4)…H(m), and corresponding spectrum value set is designated as A to each subregion respectively(2), A(3), A
(4)…A(m).
Step S3, Multivariate Analysis is used to the spectrum value set of each subregion, obtains the characteristic vector of each subregion;Fill
Divide the demand for considering actual production, select dimensionality reduction dimension of the corresponding characteristic vector quantity as the subregion;
Step S4, the subregion where sample to be tested is judged using form and aspect cluster, using the characteristic vector of the subregion to be tested
Sample carries out spectrum dimension-reduction treatment;Spectral Reconstruction, the sample to be tested after being reconstructed are carried out finally by subregion where searching
Spectral reflectivity.
A kind of 2. spectrum dimension reduction method based on form and aspect cluster according to claim 1, it is characterised in that:
Wherein, step S1 comprises the following steps:
Step S1-1, main form and aspect are determined according to hue circle principle according to training sample color characteristics, is clustered and determined using form and aspect
Training sample corresponding to each main form and aspect;
Step S1-2, the number of partitions of gained is m, and now each subregion correspond to a spectral value with same color feature
Set;
Step S1-3, calculate hue angle and saturation degree possessed by training sample;Specific method has two methods:
Method 1:Under fixed environment of observation, the CIE XYZ values corresponding to training sample spectral reflectivity are calculated, i.e.,:
Wherein,kTo adjust factor,For spectral reflectivity corresponding to training sample,For the relative spectral power distribution of spectrum
Curve,It is CIE color matching functions.
The CIE XYZ values of training sample are transformed into by CIE Lab values, hue angle h and saturation degree by color space conversion function
C, i.e.,:
Wherein,For the CIE XYZ values of training sample,It is irradiated to for CIE standard illuminants completely unrestrained anti-
The tristimulus values on beam surface;
Method 2:Spectral value collection corresponding to training sample set is carried out to use Multivariate Analysis(Such as PCA, solely
Vertical PCA, non-negative PCA etc.), obtain characteristic vector;According to the variance contribution ratio of characteristic vector(It is i.e. square
Poor contribution rate represents the ratio of the information content contained under this feature vector)Mode from big to small arranged, be designated as pc
(1), pc(2), pc(3)…pc(n).
Wherein,Be itsiIndividual characteristic vector,It is characteristic vectoriCorresponding coefficient, n are spectrum sample intervals.
Take first three characteristic vector pc(1), pc(2), pc(3), the hue angle h and saturation degree C of training sample are calculated, i.e.,:
。
A kind of 3. spectrum dimension reduction method based on form and aspect cluster according to claim 1, it is characterised in that:
Wherein, step S4 comprises the following steps:
Step S4-1, select the method for obtaining hue angle and saturation degree with calculating training sample consistent, calculate sample to be tested
Possessed hue angle and saturation degree;
Step S4-2, judge the subregion where sample to be tested, the subregion where sample to be tested is judged using form and aspect cluster, and
Subregion where treating test sample is marked;If sample to be tested just exists in the line of demarcation of two subregions, sample to be tested
Two subregions all carry out dimensionality reduction, calculate the color between the numerical value after dimensionality reduction in the case of converting various light sources and sample to be tested
Difference, the subregion where selecting multiple light courcess value of chromatism sum minimum mark place subregion as the subregion where sample to be tested.
Step S4-3, treat test sample using the characteristic vector of the subregion and carry out spectrum dimensionality reduction;Finally by where searching points
Area carries out Spectral Reconstruction, the sample spectral reflectivity to be tested after being reconstructed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711221418.5A CN107766896B (en) | 2017-11-28 | 2017-11-28 | Spectrum dimensionality reduction method based on hue clustering |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711221418.5A CN107766896B (en) | 2017-11-28 | 2017-11-28 | Spectrum dimensionality reduction method based on hue clustering |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107766896A true CN107766896A (en) | 2018-03-06 |
CN107766896B CN107766896B (en) | 2021-01-29 |
Family
ID=61276814
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711221418.5A Active CN107766896B (en) | 2017-11-28 | 2017-11-28 | Spectrum dimensionality reduction method based on hue clustering |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107766896B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109508724A (en) * | 2019-01-16 | 2019-03-22 | 北京印刷学院 | With the method for clustering classification observer's color matching functions |
CN109655155A (en) * | 2018-12-05 | 2019-04-19 | 北京印刷学院 | Lighting source colour rendering evaluation method and device |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105023239A (en) * | 2015-08-18 | 2015-11-04 | 西安电子科技大学 | Hyperspectral data dimensionality reduction method based on ultra-pixel and maximum boundary distribution |
CN105069234A (en) * | 2015-08-13 | 2015-11-18 | 武汉大学 | Spectrum dimensionality reduction method and system based on visual perception feature |
CN105869161A (en) * | 2016-03-28 | 2016-08-17 | 西安电子科技大学 | Method for selecting wave bands of hyperspectral image based on image quality assessment |
CN106408619A (en) * | 2016-09-13 | 2017-02-15 | 齐鲁工业大学 | Method of realizing cross-media color reproduction based on spectral domain |
-
2017
- 2017-11-28 CN CN201711221418.5A patent/CN107766896B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105069234A (en) * | 2015-08-13 | 2015-11-18 | 武汉大学 | Spectrum dimensionality reduction method and system based on visual perception feature |
CN105023239A (en) * | 2015-08-18 | 2015-11-04 | 西安电子科技大学 | Hyperspectral data dimensionality reduction method based on ultra-pixel and maximum boundary distribution |
CN105869161A (en) * | 2016-03-28 | 2016-08-17 | 西安电子科技大学 | Method for selecting wave bands of hyperspectral image based on image quality assessment |
CN106408619A (en) * | 2016-09-13 | 2017-02-15 | 齐鲁工业大学 | Method of realizing cross-media color reproduction based on spectral domain |
Non-Patent Citations (2)
Title |
---|
YICONG ZHOU: ""Dimension Reduction Using Spatial and Spectral Regularized Local Discriminant Embedding for Hyperspectral Image Classification"", 《IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING》 * |
刘攀: ""面向颜色再现的光谱降维方法研究"", 《包装工程》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109655155A (en) * | 2018-12-05 | 2019-04-19 | 北京印刷学院 | Lighting source colour rendering evaluation method and device |
CN109655155B (en) * | 2018-12-05 | 2020-09-29 | 北京印刷学院 | Method and device for evaluating color rendering of illumination light source |
CN109508724A (en) * | 2019-01-16 | 2019-03-22 | 北京印刷学院 | With the method for clustering classification observer's color matching functions |
CN109508724B (en) * | 2019-01-16 | 2023-10-20 | 北京印刷学院 | Method for classifying observer color matching function by using cluster analysis |
Also Published As
Publication number | Publication date |
---|---|
CN107766896B (en) | 2021-01-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Gramazio et al. | Colorgorical: Creating discriminable and preferable color palettes for information visualization | |
Sangwine et al. | The colour image processing handbook | |
Durmus | CIELAB color space boundaries under theoretical spectra and 99 test color samples | |
CN110046673A (en) | No reference tone mapping graph image quality evaluation method based on multi-feature fusion | |
CN105069234B (en) | The spectrum dimension reduction method and system of a kind of view-based access control model Perception Features | |
CN113673389B (en) | Painting illumination visual evaluation method related to spectrum power distribution of light source | |
Falomir et al. | A model for colour naming and comparing based on conceptual neighbourhood. An application for comparing art compositions | |
CN106408619B (en) | A method of realizing across media color reproductions based on spectral domain | |
CN101966083A (en) | Abnormal skin area computing system and computing method | |
US8412694B2 (en) | Colorblind accessible image search | |
CN107766896A (en) | A kind of spectrum dimension reduction method based on form and aspect cluster | |
Barnard | Computational color constancy: taking theory into practice | |
Royer | Tutorial: Background and guidance for using the ANSI/IES TM-30 method for evaluating light source color rendition | |
CN114067222B (en) | Urban water body remote sensing classification method and device | |
CN104092919B (en) | Chromatic adaptation transformation optimizing method and system for color digital imaging system | |
US7847986B2 (en) | Multiband color management | |
EP2866008B1 (en) | Tone type identification method | |
Lee et al. | Towards anovel perceptual color difference metric using circular processing of hue components | |
Canal et al. | Measuring the immeasurable: quantitative analyses of perceptual experiments | |
CN110335321B (en) | Color block matching degree evaluation method and system, storage medium and electronic equipment | |
Derhak et al. | Wpt (waypoint) shift manifold difference metrics for evaluation of varying observing‐condition (observer+ illuminant) metamerism and color inconstancy | |
CN105825020A (en) | Computing method of three-dimensional perceptual color gamut | |
CN112015417B (en) | Method and device for determining theme colors of application programs | |
US20080055599A1 (en) | Method of improving a color filter | |
Nicolás-Sáenz et al. | ABANICCO: A New Color Space for Multi-Label Pixel Classification and Color Segmentation |
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 | ||
TR01 | Transfer of patent right |
Effective date of registration: 20221212 Address after: 250306 Room 3115, No. 135, Ward Avenue, Ping'an Street, Changqing District, Jinan City, Shandong Province Patentee after: Shandong Jiqing Technology Service Co.,Ltd. Address before: 250353 College of printing and packaging engineering, 3501 Daxue Road, Changqing District, Jinan City, Shandong Province Patentee before: Qilu University of Technology |
|
TR01 | Transfer of patent right |