CN110263698A - A kind of image-characterization methods and device - Google Patents
A kind of image-characterization methods and device Download PDFInfo
- Publication number
- CN110263698A CN110263698A CN201910522872.7A CN201910522872A CN110263698A CN 110263698 A CN110263698 A CN 110263698A CN 201910522872 A CN201910522872 A CN 201910522872A CN 110263698 A CN110263698 A CN 110263698A
- Authority
- CN
- China
- Prior art keywords
- pixel
- image
- log
- value
- slope
- 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
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of image-characterization methods and device, comprising: obtains image to be processed；Calculate the linear change values of image slices vegetarian refreshments to be processed；It determines slope corresponding to central pixel point, and draws corresponding first log-log graph of the central pixel point；Obtain the corresponding first slope of each pixel in image to be processed；Acquired first slope is subjected to binary conversion treatment, obtains bianry image；Obtain the second log-log graph corresponding to the corresponding Fractal Dimension of bianry image and Fractal Dimension；Determine the second slope corresponding to each pixel, and as the corresponding fractal characteristic of the pixel.Using the embodiment of the present invention, it is intended to analyze by more Fractal Dimensions and calculates the singularity of the facial image local neighborhood information with illumination in image overall range, obtain fractal characteristic corresponding to the global regularity of distribution, final face characterization significantly shields illumination noise while being effectively retained face characteristic, provides input quantity to carry out recognition of face.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to a kind of image-characterization methods and device.
Background technique
Under complex illumination scene, the characteristic information in facial image comprising face itself and the interference from illumination
Information expresses the product that facial image i.e. facial image is reflecting component and luminance component, brightness according to illumination reflection model
Component embodies face features, and reflecting component is only related to illumination, so preferably output face characterization should be isolated
Reflecting component and shield luminance component.It is generally acknowledged that face characteristic frequency is higher, subrange changes greatly, another aspect illumination
Frequency is lower, subrange variation is smaller, it is possible to obtain image high-frequency information or small scale by the means such as filtering, decomposing
Feature carrys out approximate evaluation reflecting component, and the face as output characterizes.However existing algorithm is mostly from the angle of global or local
It goes to obtain characteristics of image, does not combine image local and global characteristics very well, cause final face characterization not ideal enough.
Summary of the invention
In view of the foregoing deficiencies of prior art, the purpose of the present invention is to provide a kind of image-characterization methods and dresses
It sets, it is intended to be analyzed the singularity of the facial image local neighborhood information with illumination by more Fractal Dimensions in image overall range
In calculated, obtain fractal characteristic corresponding to the global regularity of distribution, final face characterization can be effectively retained face characteristic
While significantly shield illumination noise, for carry out recognition of face input quantity is provided.
In order to achieve the above objects and other related objects, the present invention provides a kind of image-characterization methods, which comprises
Obtain image to be processed；
Calculate the linear change values of each pixel in the image to be processed；
The pixel centered on each of the image to be processed pixel executes step:
Multiple square windows are determined based on central pixel point, calculate pixel in each square window according to linear change values
Maximum value, minimum value and the average value of point；Based on maximum value corresponding to each square window, minimum value and average value, meter
Calculate the Hǒlder exponent of the central pixel point；According to the multiple square window, the Hǒlder exponent, the center pixel is drawn
Corresponding first log-log graph of point；According to first log-log graph, linear regression line, determine corresponding to the central pixel point
Slope；
Obtain the corresponding first slope of each pixel in the image to be processed；
According to preset value range, acquired first slope is subjected to binary conversion treatment, obtains bianry image, wherein
When any one slope in first slope is in default value range, the corresponding pixel of the slope be 1, otherwise for
0；
According to box-counting grid, second pair is obtained corresponding to the corresponding Fractal Dimension of bianry image and the Fractal Dimension
Logarithmic chart；
According to second log-log graph and linear regression line, the second slope corresponding to each pixel is determined, and
As the corresponding fractal characteristic of the pixel.
In a kind of implementation of the invention, the linear change for calculating each pixel in the image to be processed
The step of value, comprising:
Gray level image is converted by the image to be processed；
Obtain maximum gradation value, minimum gradation value in the image to be processed；
According to the pixel value of each pixel, maximum gradation value, minimum gradation value, the linear of each pixel is calculated
Change values.
It is specific public used by the linear change values for calculating each pixel in a kind of implementation of the invention
Formula expression are as follows:
Wherein, ImtnFor minimum gradation value, ImtxFor maximum gradation value, I (x, y) is the picture for the pixel that coordinate is (x, y)
Element value, I'(x, y) it is linear change values corresponding to I (x, y).
In a kind of implementation of the invention, the linear change values of basis calculate pixel in each square window
Formula used by maximum value, minimum value and average value are as follows:
Wherein, Ω is the square window of setting, μmax(x, y) is the maximum value of the linear change values of pixel, μ in window Ωmin
(x, y) is the minimum value and μ of the linear change values of pixel in window Ωmean(x, y) is the flat of the linear change values of pixel in window Ω
Mean value, Ω are the local window of setting, and the side length that ε is window Ω is ε, Ω*For all non-zero pixels points in window Ω, window
Mouth central point is (x, y), and g (s, t) is the pixel value of pixel st, and n is non-zero pixels number in Ω window.
In a kind of implementation of the invention, formula used by the Hǒlder exponent for calculating the central pixel point
Are as follows:
Wherein, hμ(x, y) is Hǒlder exponent corresponding to pixel of the coordinate for (x, y), μ (Wε(x, y)) it is measurement
Function, wherein the measurement functions are as follows: μmax(x, y), μmin(x, y) and μmean(x, y), WεThe local window for being ε for side length.
In a kind of implementation of the invention, the step of described drafting central pixel point corresponding first log-log graph,
Include:
The square window of multiple different side lengths is set；
According to Hǒlder exponent and Hǒlder exponent institute corresponding to set square window acquisition pixel
Corresponding first log-log graph.
In a kind of implementation of the invention, the step of obtaining the first log-log graph corresponding to the Hǒlder exponent,
Include:
For the central pixel point, one group is obtained by log (μ (Wε(x, y))) and the point that constitutes of log (ε), and based on being obtained
The first log-log graph of point-rendering, wherein ε=2k+1 (k=1,2,3 ..., m), m be window number and window it is maximum
The restrictive condition of side length.
It is described according to preset value range in a kind of implementation of the invention, acquired first slope is carried out two
The step of value processing, acquisition bianry image, comprising:
The first slope is divided into first quantity domain according to from the sequence of small arrival；
The previous value and next value in each domain are set；
According to the previous value and next value, binary conversion treatment is carried out to the slope in each domain；
Obtain the first quantity binary map.
In a kind of implementation of the invention, formula used by the corresponding Fractal Dimension of the acquisition bianry image is specific
Expression are as follows:
Wherein, FD is Fractal Dimension, ε '=1,2,4 ..., 16；
The step of obtaining the second log-log graph corresponding to the Fractal Dimension, comprising:
One group is obtained by log (Nε′) and-log (ε ') constitute point, and draw the second log-log graph.
In addition, the invention also discloses a kind of characterization image device, described device includes:
First obtains module, for obtaining image to be processed；
Computing module, for calculating the linear change values of each pixel in the image to be processed；
Processing module is executed: being based on for the pixel centered on each of the image to be processed pixel
Central pixel point determines multiple square windows, according to linear change values calculate the maximum value of pixel in each square window,
Minimum value and average value；Based on maximum value corresponding to each square window, minimum value and average value, the center pixel is calculated
The Hǒlder exponent of point；According to the multiple square window, the Hǒlder exponent, the central pixel point corresponding first is drawn
Log-log graph；According to first log-log graph, linear regression line, slope corresponding to the central pixel point is determined；
Second obtains module, for obtaining the corresponding first slope of each pixel in the image to be processed；
Module is obtained, for acquired first slope being carried out binary conversion treatment, obtains two according to preset value range
It is worth image, wherein when any one slope in first slope is in default value range, the corresponding pixel of the slope
Point is 1, is otherwise 0；
Third obtains module, for according to box-counting grid, obtaining the corresponding Fractal Dimension of bianry image and described dividing shape
Second log-log graph corresponding to dimension；
Determining module, for determining corresponding to each pixel according to second log-log graph and linear regression line
The second slope, and as the corresponding fractal characteristic of the pixel.
As described above, a kind of image-characterization methods provided in an embodiment of the present invention and device, first slope value is by the center
Data obtain in pixel local window, subrange variation lesser feature lower according to illumination frequency, in calculation window most
Big value, minimum value and average value, since window is smaller, then illumination component should remain unchanged, desired light when calculating first slope
It is 0 according to component slopes, maskable illumination, in addition, the image from global angle, illumination component has similitude, makes full use of
More points of shape features investigate the global regularity of distribution of Pixel Information in local window, can retain the face characteristic changed greatly and remove
The relatively slow illumination noise of variation.Therefore, by being intended to the facial image local neighborhood with illumination through more fractal analysis methods
Information, the i.e. singularity of first slope are calculated in image overall range, i.e. the second slope, obtain the global regularity of distribution,
Final face characterization significantly shields illumination noise while being effectively retained face characteristic, provides preferably for subsequent recognition of face
Input.
Detailed description of the invention
Fig. 1 is a kind of a kind of flow diagram of image-characterization methods of the embodiment of the present invention.
Fig. 2 is a kind of the first concrete application schematic diagram of image-characterization methods of the embodiment of the present invention.
Fig. 3 is a kind of second of concrete application schematic diagram of image-characterization methods of the embodiment of the present invention.
Fig. 4 is a kind of the third concrete application schematic diagram of image-characterization methods of the embodiment of the present invention.
Specific embodiment
Illustrate embodiments of the present invention below by way of specific specific example, those skilled in the art can be by this specification
Other advantages and efficacy of the present invention can be easily understood for disclosed content.The present invention can also pass through in addition different specific realities
The mode of applying is embodied or practiced, the various details in this specification can also based on different viewpoints and application, without departing from
Various modifications or alterations are carried out under spirit of the invention.
Please refer to Fig. 1-4.It should be noted that only the invention is illustrated in a schematic way for diagram provided in the present embodiment
Basic conception, only shown in schema then with related component in the present invention rather than component count, shape when according to actual implementation
Shape and size are drawn, when actual implementation kenel, quantity and the ratio of each component can arbitrarily change for one kind, and its component cloth
Office's kenel may also be increasingly complex.
As shown in Figure 1, the embodiment of the present invention provides a kind of image-characterization methods, which comprises
S101 obtains image to be processed.
Image to be processed in the embodiment of the present invention is the image comprising face, to carry out at further facial image
Reason.
S102 calculates the linear change values of each pixel in the image to be processed.
Gray value is carried out to image first after input picture and draws high operation, tentatively to mitigate the influence of extreme path photograph.
The minimum gradation value I of input picture is calculated firstmin=min [I (x, y)] and maximum gradation value Imax=max [I (x, y)], will
IminAnd ImaxIt is respectively mapped to 0 and 255, then the value of each pixel can be changed linearly are as follows:
S103, the pixel centered on each of the image to be processed pixel execute step:
Multiple square windows are determined based on central pixel point, calculate pixel in each square window according to linear change values
Maximum value, minimum value and the average value of point；Based on maximum value corresponding to each square window, minimum value and average value, meter
Calculate the Hǒlder exponent of the central pixel point；According to the multiple square window, the Hǒlder exponent, the center pixel is drawn
Corresponding first log-log graph of point；According to first log-log graph, linear regression line, determine corresponding to the central pixel point
Slope.
For linear change values I'(x, y corresponding to facial image) in pixel (x, y), the rectangular of a certain size is set
Window observes the rule of the grey scale pixel value in the window, which can be acquired with measurement functions μ (), can be used
Different μ () functions acquires the measured value of pixel in different windows, and the present invention uses μmax(x, y), μmin(x, y) and
μmean(x, y) acquires the maximum value of pixel, minimum value and pixel average in window.It is defined respectively as:
Ω is the local window of setting, and its side length is ε, Ω*For all non-zero pixels points in Ω window.Window center
Point is (x, y), and g (s, t) is the pixel value of pixel (s, t), and n is non-zero pixels number in Ω window.
It calculates the α feature of pixel (x, y), i.e. first slope: calculating the Hall moral of each pixel of facial image firstIndex hμ(x, y), calculation formula are as follows:
Wherein, μ () function is above-mentioned described measurement functions, can select different measurement methods, WεFor above-mentioned institute
The side length of description is the local window of ε.
Simplify the above calculating by drawing log-log graph, be arranged different ε value ε=2k+1 (k=1,2,3 ...,
M), the first slope α for the linear regression line put by this group is calculated as a result, as pixel α feature, as shown in Figure 2.
S104 obtains the corresponding first slope of each pixel in the image to be processed.
It repeats the above steps for each pixel, obtains the corresponding characteristic pattern of first slope α of whole picture facial image.
It it is understood that being directed to first slope, is obtained by data in central pixel point local window, according to illumination frequency
Rate is lower, subrange changes lesser feature, maximum value, minimum value and average value in calculation window, since window is smaller, then
Illumination component should remain unchanged, and desired light is 0 according to component slopes when calculating first slope, maskable illumination.
Acquired first slope is carried out binary conversion treatment, obtains bianry image by S105 according to preset value range,
Wherein, when any one slope in first slope is in default value range, the corresponding pixel of the slope is 1,
It otherwise is 0.
All first slope α values are divided into R domain, α from small to larger(x, y) indicates the institute in these domains in r-th of domain
There is α value, by the lower bound α that each domain is arrangedrLowWith upper bound αrUpValue obtain binary imageThe binary image and
Original input image dimension is consistent, and calculation method is as follows:
Thus it is possible to obtain R bianry image.
S106 is obtained corresponding to the corresponding Fractal Dimension of bianry image and the Fractal Dimension according to box-counting grid
Second log-log graph.
For a width bianry imageIts Fractal Dimension FD is calculated using box-counting (box-counting) method, two
It is worth on image and covers one layer of grid, the side length of lattice is ε ' in grid, then calculates the number N of non-empty latticeε′.In binary map
As the grid of upper covering different scale, the side length of lattice is ε ' in grid, and the value of different ε ' is arranged, obtains different binary maps
As the quantity of upper non-zero lattice.Then the Fractal Dimension of the bianry image can be calculated, calculation formula is as follows:
Simplify the above calculating by drawing log-log graph, ε '=1 is set, 2,4 ..., 16, thus to obtain one group by log
(Nε′) and the point that constitutes of-log (ε ') and be plotted on log-log graph, calculate the linear regression line put by this group second is oblique
Rate f (α), as the fractal characteristic of the bianry image, schematic diagram corresponding to the second slope f (α) as shown in Figure 3.
S107 is determined corresponding to each pixel second tiltedly according to second log-log graph and linear regression line
Rate, and as the corresponding fractal characteristic of the pixel.
It, then can be by one group of log (N according to the quantity for obtaining non-zero lattice on different bianry imagesε′) and-log
The second log-log graph of point-rendering that (ε ') is constituted calculates the second slope, more fractal feature representations as the input picture.
Until more fractal characteristic f (α) of all bianry images are all calculated, one group then is obtained by different bianry images
Corresponding fractal characteristic constitutes the more points of shapes spectrum of former input facial image.
Each second slope f (α) corresponds to a αrLowValue, with f (α) and αrLowA curve is drawn to indicate f (α) and α
Relationship, i.e., more fractal feature representations.F (α) characteristic pattern is calculated by first slope α characteristic pattern correspondence, as shown in figure 4, by f
(α) characteristic pattern characterizes output as the facial image under complex illumination scene.
On the basis of being based on first slope, the second slope, that is, the image from global angle, illumination are calculated again
Component has similitude, makes full use of more points of shape features to investigate the global regularity of distribution of Pixel Information in local window, can retain
The face characteristic that changes greatly and remove the relatively slow illumination noise of variation.
In addition, the embodiment of the invention also discloses a kind of characterization image device, described device includes:
First obtains module, for obtaining image to be processed；
Computing module, for calculating the linear change values of each pixel in the image to be processed；
Processing module is executed: being based on for the pixel centered on each of the image to be processed pixel
Central pixel point determines multiple square windows, according to linear change values calculate the maximum value of pixel in each square window,
Minimum value and average value；Based on maximum value corresponding to each square window, minimum value and average value, the center pixel is calculated
The Hǒlder exponent of point；According to the multiple square window, the Hǒlder exponent, the central pixel point corresponding first is drawn
Log-log graph；According to first log-log graph, linear regression line, slope corresponding to the central pixel point is determined；
Second obtains module, for obtaining the corresponding first slope of each pixel in the image to be processed；
Module is obtained, for acquired first slope being carried out binary conversion treatment, obtains two according to preset value range
It is worth image, wherein when any one slope in first slope is in default value range, the corresponding pixel of the slope
Point is 1, is otherwise 0；
Third obtains module, for according to box-counting grid, obtaining the corresponding Fractal Dimension of bianry image and described dividing shape
Second log-log graph corresponding to dimension；
Determining module, for determining corresponding to each pixel according to second log-log graph and linear regression line
The second slope, and as the corresponding fractal characteristic of the pixel.
The above-described embodiments merely illustrate the principles and effects of the present invention, and is not intended to limit the present invention.It is any ripe
The personage for knowing this technology all without departing from the spirit and scope of the present invention, carries out modifications and changes to above-described embodiment.Cause
This, institute is complete without departing from the spirit and technical ideas disclosed in the present invention by those of ordinary skill in the art such as
At all equivalent modifications or change, should be covered by the claims of the present invention.
Claims (10)
1. a kind of image-characterization methods, which is characterized in that the described method includes:
Obtain image to be processed；
Calculate the linear change values of each pixel in the image to be processed；
The pixel centered on each of the image to be processed pixel executes step:
Multiple square windows are determined based on central pixel point, calculate pixel in each square window according to linear change values
Maximum value, minimum value and average value；Based on maximum value corresponding to each square window, minimum value and average value, calculating should
The Hǒlder exponent of central pixel point；According to the multiple square window, the Hǒlder exponent, the central pixel point pair is drawn
The first log-log graph answered；According to first log-log graph, linear regression line, determine oblique corresponding to the central pixel point
Rate；
Obtain the corresponding first slope of each pixel in the image to be processed；
According to preset value range, acquired first slope is subjected to binary conversion treatment, obtains bianry image, wherein the
When any one slope in one slope is in default value range, otherwise it is 0 that the corresponding pixel of the slope, which is 1,；
According to box-counting grid, the second double-log corresponding to the corresponding Fractal Dimension of bianry image and the Fractal Dimension is obtained
Figure；
According to second log-log graph and linear regression line, the second slope corresponding to each pixel, and conduct are determined
The corresponding fractal characteristic of the pixel.
2. image-characterization methods according to claim 1, which is characterized in that each in the calculating image to be processed
The step of linear change values of a pixel, comprising:
Gray level image is converted by the image to be processed；
Obtain maximum gradation value, minimum gradation value in the image to be processed；
According to the pixel value of each pixel, maximum gradation value, minimum gradation value, the linear change of each pixel is calculated
Value.
3. image-characterization methods according to claim 1 or 2, which is characterized in that the line for calculating each pixel
Property change values used by specific formula expression are as follows:
Wherein, IminFor minimum gradation value, ImaxFor maximum gradation value, I (x, y) is the pixel value for the pixel that coordinate is (x, y),
I ' (x, y) is linear change values corresponding to I (x, y).
4. image-characterization methods according to claim 3, which is characterized in that the linear change values of basis calculate each
Formula used by the maximum value, minimum value of pixel and average value in square window are as follows:
Wherein, Ω is the square window of setting, μmax(x, y) is the maximum value of the linear change values of pixel, μ in window Ωmin(x, y)
It is the minimum value and μ of the linear change values of pixel in window Ωmean(x, y) is the average value of the linear change values of pixel in window Ω,
Ω is the local window of setting, and the side length that ε is window Ω is ε, Ω*For all non-zero pixels points in window Ω, in window
Heart point is (x, y), and g (s, t) is the pixel value of pixel st, and n is non-zero pixels number in Ω window.
5. image-characterization methods according to claim 4, which is characterized in that the Hall moral for calculating the central pixel point
Formula used by index are as follows:
Wherein, hμ(x, y) is Hǒlder exponent corresponding to pixel of the coordinate for (x, y), μ (Wε(x, y)) it is measurement letter
Number, wherein the measurement functions are as follows: μmax(x, y), μmin(x, y) and μmean(x, y), WεThe local window for being ε for side length.
6. image-characterization methods according to claim 4 or 5, which is characterized in that described drafting central pixel point is corresponding
The first log-log graph the step of, comprising:
The square window of multiple different side lengths is set；
According to corresponding to Hǒlder exponent corresponding to set square window acquisition pixel and the Hǒlder exponent
The first log-log graph.
7. image-characterization methods according to claim 6, which is characterized in that obtain corresponding to the Hǒlder exponent
The step of one log-log graph, comprising:
For the central pixel point, one group is obtained by log (μ (Wε(x, y))) and the point that constitutes of log (ε), and based on obtained
The first log-log graph of point-rendering, wherein (k=1,2,3 ..., m), m is the number and window maximal side of window to ε=2k+1
Restrictive condition.
8. image-characterization methods according to claim 7, which is characterized in that it is described according to preset value range, it will be obtained
The step of first slope taken carries out binary conversion treatment, obtains bianry image, comprising:
The first slope is divided into the first quantity domain according to sequence from small to large；
The previous value and next value in each domain are set；
According to the previous value and next value, binary conversion treatment is carried out to the slope in each domain；
Obtain the first quantity binary map.
9. image-characterization methods according to claim 7 or 8, which is characterized in that described corresponding point of acquisition bianry image
Formula used by shape dimension embodies are as follows:
Wherein, FD is Fractal Dimension, ε '=1,2,4 ..., 16；
The step of obtaining the second log-log graph corresponding to the Fractal Dimension, comprising:
One group is obtained by log (Nε′) and-log (ε ') constitute point, and draw the second log-log graph.
10. a kind of characterization image device, which is characterized in that described device includes:
First obtains module, for obtaining image to be processed；
Computing module, for calculating the linear change values of each pixel in the image to be processed；
Processing module is executed for the pixel centered on each of the image to be processed pixel: being based on center
Pixel determines multiple square windows, and the maximum value, minimum of pixel in each square window are calculated according to linear change values
Value and average value；Based on maximum value corresponding to each square window, minimum value and average value, the central pixel point is calculated
Hǒlder exponent；According to the multiple square window, the Hǒlder exponent, draw the central pixel point it is corresponding first pair it is right
Number figure；According to first log-log graph, linear regression line, slope corresponding to the central pixel point is determined；
Second obtains module, for obtaining the corresponding first slope of each pixel in the image to be processed；
Module is obtained, for acquired first slope being carried out binary conversion treatment, obtains binary map according to preset value range
Picture, wherein when any one slope in first slope is in default value range, the corresponding pixel of the slope is
1, it is otherwise 0；
Third obtains module, for obtaining the corresponding Fractal Dimension of bianry image and the Fractal Dimension according to box-counting grid
The second corresponding log-log graph；
Determining module determines corresponding to each pixel for according to second log-log graph and linear regression line
Two slopes, and as the corresponding fractal characteristic of the pixel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910522872.7A CN110263698B (en) | 2019-06-17 | 2019-06-17 | Image characterization method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910522872.7A CN110263698B (en) | 2019-06-17 | 2019-06-17 | Image characterization method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110263698A true CN110263698A (en) | 2019-09-20 |
| CN110263698B CN110263698B (en) | 2022-09-30 |
Family
ID=67918723
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910522872.7A Active CN110263698B (en) | 2019-06-17 | 2019-06-17 | Image characterization method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110263698B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5787201A (en) * | 1996-04-09 | 1998-07-28 | The United States Of America As Represented By The Secretary Of The Navy | High order fractal feature extraction for classification of objects in images |
| CN107862267A (en) * | 2017-10-31 | 2018-03-30 | 天津科技大学 | Face recognition features' extraction algorithm based on full symmetric local weber description |
-
2019
- 2019-06-17 CN CN201910522872.7A patent/CN110263698B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5787201A (en) * | 1996-04-09 | 1998-07-28 | The United States Of America As Represented By The Secretary Of The Navy | High order fractal feature extraction for classification of objects in images |
| CN107862267A (en) * | 2017-10-31 | 2018-03-30 | 天津科技大学 | Face recognition features' extraction algorithm based on full symmetric local weber description |
Non-Patent Citations (2)
| Title |
|---|
| 张洁玉: "基于图像分块的局部阈值二值化方法", 《计算机应用》 * |
| 边后琴: "人脸图像的LTP特征提取", 《上海电力学院学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110263698B (en) | 2022-09-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108876743B (en) | Image rapid defogging method, system, terminal and storage medium | |
| CN105354865B (en) | The automatic cloud detection method of optic of multispectral remote sensing satellite image and system | |
| CN106846276B (en) | Image enhancement method and device | |
| CN108921800A (en) | Non-local mean denoising method based on form adaptive search window | |
| Lai et al. | Improved local histogram equalization with gradient-based weighting process for edge preservation | |
| CN110428371A (en) | Image defogging method, system, storage medium and electronic equipment based on super-pixel segmentation | |
| CN107464226B (en) | A kind of image de-noising method based on improvement two-dimensional empirical mode decomposition algorithm | |
| CN108537756A (en) | Single image to the fog method based on image co-registration | |
| CN104272347B (en) | Removal includes the image processing apparatus and its method of the mist in static image | |
| CN105847703A (en) | Image processing method and electronic device | |
| CN106157264B (en) | Large area image uneven illumination bearing calibration based on empirical mode decomposition | |
| CN104504662A (en) | Homomorphic filtering based image processing method and system | |
| CN103854267A (en) | Image fusion and super-resolution achievement method based on variation and fractional order differential | |
| CN106934770B (en) | A kind of method and apparatus for evaluating haze image defog effect | |
| CN111612725A (en) | Image fusion method based on contrast enhancement of visible light image | |
| CN109961408A (en) | The photon counting Image denoising algorithm filtered based on NSCT and Block- matching | |
| CN106937059A (en) | Image synthesis method and system based on Kinect | |
| CN108647597A (en) | A kind of wrist recognition methods, gesture identification method, device and electronic equipment | |
| Bhargava et al. | An adaptive method for edge preserving denoising | |
| CN111444555B (en) | Temperature measurement information display method and device and terminal equipment | |
| CN103839245A (en) | Retinex night color image enhancement method based on statistical regularities | |
| JP3406577B2 (en) | Object recognition method | |
| CN106097261A (en) | Image processing method and device | |
| CN113052923B (en) | Tone mapping method, tone mapping apparatus, electronic device, and storage medium | |
| Zeng et al. | High dynamic range infrared image compression and denoising |
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 |