CN104077744A - Image enhancing method and device - Google Patents
Image enhancing method and device Download PDFInfo
- Publication number
- CN104077744A CN104077744A CN201310103337.0A CN201310103337A CN104077744A CN 104077744 A CN104077744 A CN 104077744A CN 201310103337 A CN201310103337 A CN 201310103337A CN 104077744 A CN104077744 A CN 104077744A
- Authority
- CN
- China
- Prior art keywords
- mrow
- mfrac
- frequency domain
- mtd
- centerdot
- 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
- 238000000034 method Methods 0.000 title claims abstract description 28
- 230000002708 enhancing effect Effects 0.000 title abstract description 7
- 230000009466 transformation Effects 0.000 claims abstract description 12
- 230000001965 increasing effect Effects 0.000 claims abstract description 5
- 239000011159 matrix material Substances 0.000 claims description 20
- 230000003044 adaptive effect Effects 0.000 claims description 7
- 230000001131 transforming effect Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000000354 decomposition reaction Methods 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000003707 image sharpening Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Image Processing (AREA)
Abstract
The invention provides an image enhancing method and an image enhancing device. The image enhancing method comprises the following steps that: three components R, G and B of color images are respectively subjected to frequency domain transformation for obtaining frequency domain transformation coefficients; the frequency domain transformation coefficients with the absolute values smaller than a preset threshold are set to be zero, and in addition, the frequency domain transformation coefficients with the absolute values greater than or equal to the preset threshold are set to be self-relevant values by considering the continuity of weak detail signals of the color images and the undistortion of strong detail images; the set frequency domain transformation coefficients are subjected to extremum processing for increasing the differences; and the frequency domain transformation coefficients subjected to extremum processing are subjected to frequency domain inverse transformation to obtain three components Rn, Gn and Bn.
Description
Technical Field
The present invention relates to the field of image processing, and more particularly, to an image enhancement method and apparatus.
Background
The purpose of image enhancement is to improve the quality and visual effect of images by processing the images, or to convert the images into a form more suitable for human eye viewing or machine analysis and recognition, so as to more effectively acquire useful information from the images.
Generally, image enhancement includes several aspects of noise filtering, contrast enhancement, saturation enhancement, and image sharpening. Specifically, the image is often interfered and affected by various noises during the processes of acquisition, transmission and storage, so that the image is degraded, and noise filtering is used for reducing the influence of the noises on the image to the maximum extent; contrast enhancement is used for improving the visibility of the image and highlighting target information or hidden information; the saturation enhancement is mainly used for improving the layering sense of the image and enabling the color of the image to be more gorgeous; the image sharpening is used for enhancing the outline texture of the image and improving the definition of a target object, so that the target object is easier to detect and recognize.
Image enhancement can be divided into two types, spatial domain image enhancement and transform domain (such as frequency domain) image enhancement, depending on the processing space. Specifically, the spatial domain image enhancement mainly includes histogram equalization, histogram modification, and the like, and the transform in the transform domain image enhancement mainly employs fourier transform, wavelet transform, and the like. However, due to their own limitations or theoretical drawbacks, the above image enhancement processing is not good, such as causing image blur, losing detail contour information, and severe image distortion.
In addition, when an image is captured through a portable terminal, the image quality may be poor due to various influences of hardware or surrounding environment, such as noise, poor contrast, and image blurring, and none of the above conventional image enhancement techniques can effectively process the image.
Disclosure of Invention
According to an exemplary embodiment of the present invention, there is provided an image enhancement method including: frequency domain transforming the three components R, G, B of the color image to obtain frequency domain transform coefficients, respectively; setting the frequency domain transform coefficients having absolute values less than a predetermined threshold to zero, and setting the frequency domain transform coefficients having absolute values greater than or equal to the predetermined threshold to values related to themselves in consideration of continuity of the weak detail signal and undistorted strong detail signal of the color image; increasing the diversity by polarizing the set frequency domain transform coefficients; and performing an inverse frequency domain transform on the polarized frequency domain transform coefficients to obtain three components Rn, Gn, Bn.
The image enhancement method may further include: the three components R, G, B are normalized before the three components R, G, B of the color image are frequency domain transformed, respectively.
The frequency domain transform may be a Contourlet transform (Contourlet) transform and the frequency domain transform coefficients may be Contourlet transform coefficients.
The step of setting the frequency domain transform coefficients may comprise: the frequency domain transform coefficients are set by the following equation,
where δ represents a predetermined threshold, Coeff represents a Contourlet transform coefficient, and Coeff δ represents a settingThe following Coeff, sgn, denotes sign operation, max denotes maximum operation, mean2 denotes mean operation, std2 denotes standard deviation operation, σ denotes noise standard deviation estimate of R, G, B, which is given by equationA calculation is performed in which Median denotes the Median operation, Coeff1 denotes the first layer coefficient of Coeff, and r has a value of 0.6745.
δ may employ a bayesian shrinkage (bayessshrnk) threshold.
δ may employ an adaptive bayesian shrinkage (bayessshrnk) threshold.
The step of quantizing the set frequency domain transform coefficients may include: the first layer coefficient in the set Contourlet transform coefficients is quantized.
The step of quantizing the first layer coefficient in the set Contourlet transform coefficients may include: selecting a two-dimensional area for each of a plurality of subgraphs constituted by coefficients in the same direction among the coefficients of the first layer converted into two dimensions of Coeff δ by a predetermined matrix; performing convolution processing on each selected two-dimensional area through a mean matrix to obtain a convolution value; for each selected two-dimensional region, the value of the center point takes the maximum value in the selected two-dimensional region when the value of the center point of the selected two-dimensional region is greater than or equal to the convolution value, and the value of the center point takes the minimum value in the selected two-dimensional region when the value of the center point of the selected two-dimensional region is less than the convolution value.
The image enhancement method may further include: spatial domain image enhancement is performed by using the three components Rn, Gn, Bn.
The step of spatial domain image enhancement by using the three components Rn, Gn, Bn may comprise: convolution processing is respectively carried out on the three components Rn, Gn and Bn through a matrix in a pyramid form to obtain RnF, GnF and BnF, the three components RNS, GNS and BNS of the spatial domain image enhancement are obtained through the following equations,
where Rn (i, j), Gn (i, j), and Bn (i, j) respectively represent values at coordinates (i, j) in Rn, Gn, Bn, RnF (i, j), GnF (i, j), BnF (i, j) respectively represent values at coordinates (i, j) in RnF, GnF, BnF, RNS, GNS (i, j), BNS (i, j) respectively represent values at coordinates (i, j) in RNS, GNS, BNS, t represents an overall saturation adjustment factor, ζ represents a local contrast adjustment factor, and RNS (i, j), GNS (i, j), BNS (i, j) are limited to corresponding end point values of a predetermined range when RNS (i, j), GNS (i, j), BNS (i, j) exceed the predetermined range.
The color image can be obtained by a portable terminal.
According to an exemplary embodiment of the present invention, there is provided an image enhancement apparatus including: a frequency domain transform unit that frequency domain transforms the three components R, G, B of the color image, respectively, to obtain frequency domain transform coefficients; a threshold processing unit that sets the frequency domain transform coefficient whose absolute value is smaller than a predetermined threshold to zero, and sets the frequency domain transform coefficient whose absolute value is greater than or equal to the predetermined threshold to a value related to itself in consideration of continuity of the color image weak detail signal and undistorted strong detail signal; a maximum and minimum processing unit, which increases the difference by polarizing the set frequency domain transformation coefficient; and a frequency domain inverse transform unit that performs frequency domain inverse transform on the polarized frequency domain transform coefficients to obtain three components Rn, Gn, Bn.
The image enhancement device may further include: and a spatial domain image enhancement unit for performing spatial domain image enhancement by using the three components Rn, Gn, Bn.
The color image can be obtained by a portable terminal.
Additional aspects and/or advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and other objects and features of the present invention will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings, in which:
fig. 1 is a flowchart illustrating an image enhancement method according to an exemplary embodiment of the present invention;
FIG. 2 is a flowchart illustrating the extremization step in FIG. 1 according to an exemplary embodiment of the present invention;
fig. 3 is a block diagram illustrating an image enhancement apparatus according to an exemplary embodiment of the present invention.
Detailed Description
Reference will now be made in detail to the exemplary embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout.
Fig. 1 is a flowchart illustrating an image enhancement method according to an exemplary embodiment of the present invention.
Referring to fig. 1, in step S110, RGB spatial decomposition may be performed on a color image I (a three-dimensional matrix m × n × 3, where m and n are positive integers) to obtain three components R, G and B, each of which is a two-dimensional matrix R, G and B. Specifically, the RGB spatial decomposition may be performed according to the following equation 1.
Wherein, R (i, j), G (i, j) and B (i, j) are integers, which respectively represent values on coordinates (i, j) in the two-dimensional matrixes R, G and B, and the value ranges are all [0, 255], i is an integer, the value range is [0, m-1], j is an integer, and the value range is [0, n-1 ].
In step S120, the normalization process may be performed on the three components R, G, B, respectively. Specifically, each element in the three components R, G, B may be divided by a predetermined value for normalization, where the predetermined value may be, but is not limited to, 255.
It should be understood that the above steps S110, S120 are not necessary steps, such as in the case where the three components R, G, B of the color image are known in advance, the above step S110 may be omitted, and the step S120 performed only for the convenience of calculation may be omitted by those skilled in the art according to actual needs.
In step S130, the three components R, G, B of the color image are respectively frequency-domain transformed to obtain frequency-domain transform coefficients. Here, for example only, the frequency domain transform may be a Contourlet (Contourlet) transform, which is a Contourlet transform coefficient, and accordingly, the Contourlet transform coefficients obtained by performing the Contourlet transform on the three components R, G, B of the color image may be Rcoeff, Gcoeff, and Bcoeff, respectively, which are one-dimensional vectors. Hereinafter, for convenience of description, the frequency domain transform is described as a Contourlet transform, but it should be understood that the invention may also be used with other frequency domain transforms, such as wavelet transforms, non-sampled Contourlet transforms (NSCTs), or other multi-resolution geometric analysis tools.
In step S140, the Contourlet transform coefficient having an absolute value less than the predetermined threshold is set to zero, and the Contourlet transform coefficient having an absolute value greater than or equal to the predetermined threshold is set to a value related to itself in consideration of the continuity of the color image weak detail signal and the strong detail signal non-distortion. Specifically, the Contourlet transform coefficient may be set by equation 2 below.
Where δ denotes a predetermined threshold value, Coeff (i.e., a unified expression of Rcoeff, Gcoeff, and Bcoeff) denotes a Contourlet transform coefficient, Coeff δ (i.e., a unified expression of Rcoeff δ, Gcoeff δ, and Bcoeff δ) denotes Coeff after setting, sgn denotes a sign-taking operation, max denotes a maximum-value-taking operation, mean2 denotes a mean-value-taking operation, std2 denotes a standard-difference-taking operation, and σ denotes a noise standard-variance estimate of R, G, B, the value of which is calculated by the equationA calculation is performed in which Median denotes the Median operation, Coeff1 denotes the first layer coefficient of Coeff, and r has a value of 0.6745.
It should be understood that equation 2 above is only an example of setting the Contourlet transform coefficients, and those skilled in the art can set the Contourlet transform coefficients by other schemes, such as a hard threshold function, a soft threshold function, a compromise threshold function, a semi-soft threshold function, etc., as shown in the following equations:
wherein, δ1、δ2Respectively representing a threshold value, w representing a Contourlet coefficient, wδRepresents the set Contourlet coefficient, and the value range of gamma is [0, 1]]。
Further, for the predetermined threshold δ in equation 2, a bayesian shrinkage (bayesian shrnk) threshold or an adaptive bayesian shrnk threshold may be employed. The equation for the Bayesian spring threshold isWherein,is a noise standard deviation estimate, σ, of R, G, BsA standard deviation estimate of R, G, B; the equation for the adaptive Bayes spring threshold isWherein,threshold values representing the layer 1 j direction of Contourlet transform estimated by Bayes Shrink threshold values, whose values are given by the equationAnd performing calculation, wherein the Median represents the Median operation,the high-frequency coefficient in the j direction of the layer 1 of Contourlet transform is represented, and the value of r is 0.6745;denotes the adaptive threshold for the layer 1J direction, J1Indicates the total number of directions of the layer 1,representing the energy value in the J direction of layer 1, η is typically min (J)1) And min represents the minimum value operation.
The foregoing bayesian spring threshold and the adaptive bayesian spring threshold are both thresholds in the prior art, and therefore will not be described in detail here. Further, it should be appreciated that in addition to using a Bayesian Shrink threshold or an adaptive Bayesian Shrink threshold, one skilled in the art can employ other thresholds such as a uniform (Visuscrink) threshold, a confidence interval threshold based on a zero-mean normal distribution, etc., as desired.
In step S150, the difference is increased by quantizing the set Contourlet transform coefficients. The step of quantizing the set Contourlet transform coefficients will be described in more detail below with reference to fig. 2.
Fig. 2 is a flowchart illustrating the extremization step in fig. 1 according to an exemplary embodiment of the present invention.
Referring to fig. 2, in step S151, a two-dimensional area centered at each point in each of a plurality of subgraphs constituted by coefficients of Coeff δ converted to the same direction in the two-dimensional first-layer coefficients is selected by a predetermined matrix.
In step S152, each selected two-dimensional region is convolved by a mean matrix to obtain a convolution value, wherein the mean matrix may be a two-dimensional matrix in odd-by-odd form, elements therein may have the same value, and the sum of all elements may be 1.
In step S153, for each selected two-dimensional region, when the value of the center point of the selected two-dimensional region is greater than or equal to the convolution value, the value of the center point takes the maximum value in the selected two-dimensional region, and when the value of the center point of the selected two-dimensional region is less than the convolution value, the value of the center point takes the minimum value in the selected two-dimensional region.
For example, suppose a subgraph consisting of coefficients in the same direction in the first layer coefficients transformed into two dimensions of Rcoeff δ is The predetermined matrix is a 3 x 3 form matrix and if 0.6 is selected for the point at position (3, 3) then the two-dimensional area is selected to be If 0.1 is for the point at position (1, 1), then the two-dimensional region is selected to be And so on. Assume a mean matrix of The two-dimensional area of point 0.6 at said position (3, 3) is convolved 0.2 x 1/9+0.2 x 1/9+0.2 x 1/9+0.3 x 1/9+0.3 x 1/9+0.4 x 1/9+0.4 x 1/9+0.4 x 1/9+0.6 x 1/9 to obtain the convolution value 1/3. Since the value of the center point of the two-dimensional area 0.6 is greater than 1/3, the value of the center point takes the maximum value in the selected two-dimensional area, i.e., 0.6.
It should be understood that the above-described extremization process is only exemplary, and those skilled in the art can perform extremization process of other algorithms or schemes according to actual needs.
Referring back to fig. 1, in step S160, the polarized Contourlet transform coefficients are subjected to a Contourlet inverse transform to obtain three components Rn, Gn, Bn, which are each a two-dimensional matrix.
In step S170, spatial domain image enhancement may be performed by using the three components Rn, Gn, Bn. Specifically, the three components Rn, Gn, Bn may be respectively convolved by a matrix in a pyramid form to obtain three components RnF, GnF, BnF, and then three components RNS, GNS, BNS for spatial domain image enhancement may be obtained by the following equation 3, wherein the matrix in the pyramid form may select values of elements in a pyramid form, that is, a value in the middle of the matrix is the largest and a value farther from the center is smaller, and the sum of all elements may be 1, such as,
wherein Rn (i, j), Gn (i, j), Bn (i, j) respectively represent values at coordinates (i, j) in the two-dimensional matrices Rn, Gn, Bn; accordingly, RnF (i, j), GnF (i, j), BnF (i, j) respectively represent values at coordinates (i, j) in the two-dimensional matrices RnF, GnF, BnF; RNS, GNS and BNS respectively represent three components Rn, Gn and Bn after spatial domain image enhancement, and RNS (i, j), GNS (i, j) and BNS (i, j) respectively represent values on coordinates (i, j) in two-dimensional matrixes RNS, GNS and BNS; t represents an overall saturation adjustment factor, which can be generally a small non-negative number, such as 0.1, according to actual needs; zeta represents the local contrast regulating factor, the value range is [0, 1], when setting it, it can be adjusted by judging the similarity (or difference) between the central point and the adjacent area of the color image to realize the local enhancement of the color image, and for the special image or special requirement, it can adjust zeta to control the local enhancement of the color image, and ensure the color image not to distort; when the RNS (i, j), GNS (i, j), BNS (i, j) are outside the predetermined range, the RNS (i, j), GNS (i, j), BNS (i, j) may be limited to the respective endpoints of the predetermined range.
It should be understood that the above-mentioned spatial domain image enhancement processing is only exemplary, and those skilled in the art can perform spatial domain image enhancement processing of other algorithms or schemes, such as saturation boosting, contrast enhancement, etc., according to actual needs.
Further, if the steps S110 and S120 are performed previously, the RNS, GNS, BNS may be restored to the color image Inew (three-dimensional matrix m × n × 3, where m, n are positive integers) at step S180. Specifically, the color image Inew can be restored according to the following equation 4.
Wherein i is an integer and has a value range of [0, m-1], j is an integer and has a value range of [0, n-1 ].
Fig. 3 is a block diagram illustrating an image enhancement apparatus according to an exemplary embodiment of the present invention.
Referring to fig. 3, the image enhancing apparatus 300 according to an exemplary embodiment of the present invention may include a frequency domain transforming unit 310, a threshold processing unit 320, a minimum maximum processing unit 330, and a frequency domain inverse transforming unit 340.
Frequency-domain transform unit 310 may perform a frequency-domain transform on each of the three components R, G, B of the color image to obtain frequency-domain transform coefficients.
The threshold processing unit 320 may set the frequency domain transform coefficient having an absolute value smaller than a predetermined threshold to zero, and set the frequency domain transform coefficient having an absolute value greater than or equal to the predetermined threshold to a value related to itself in consideration of continuity of the color image weak detail signal and undistorted strong detail signal.
The minimum-maximum processing unit 330 may increase the diversity by polarizing the set frequency domain transform coefficients.
The frequency-domain inverse transform unit 340 may perform a frequency-domain inverse transform on the polarized frequency-domain transform coefficients to obtain three components Rn, Gn, Bn.
Furthermore, the image enhancement apparatus 300 may further optionally include a spatial domain image enhancement unit 350 that may perform spatial domain image enhancement by using the three components Rn, Gn, Bn; the image enhancement apparatus 300 further optionally comprises an RGB spatial decomposition unit, a normalization unit and a color image restoration unit to perform RGB spatial decomposition of the color image, normalization of the three components R, G, B and restoration of the color image, respectively; the frequency domain transforming unit 310, the threshold processing unit 320, the minimum-maximum processing unit 330, and the frequency domain inverse transforming unit 340 may also be configured as a single frequency domain image enhancing unit to achieve the same function.
According to an exemplary embodiment of the present invention, in the color image frequency domain enhancement process, a threshold setting scheme having adaptivity and universality is proposed in consideration of the influence of a color image, noise, and the frequency domain itself, thereby effectively removing the noise of the color image; by increasing the difference of the outermost transform coefficients, the edge difference of the image is improved, and particularly the streak effect generated by the Contourlet transform is reduced. In the color image space domain enhancement processing, the definition, the contrast and the vividness of the color image are improved according to the difference between the central point and the neighborhood of the color image. Furthermore, the present invention is particularly applicable to images of poor quality taken by portable terminals (such as mobile phones, tablets, portable digital assistants, etc.).
While the invention has been shown and described with reference to certain exemplary embodiments, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims and their equivalents.
Claims (14)
1. An image enhancement method, comprising:
frequency domain transforming the three components R, G, B of the color image to obtain frequency domain transform coefficients, respectively;
setting the frequency domain transform coefficients having absolute values less than a predetermined threshold to zero, and setting the frequency domain transform coefficients having absolute values greater than or equal to the predetermined threshold to values related to themselves in consideration of continuity of the weak detail signal and undistorted strong detail signal of the color image;
increasing the diversity by polarizing the set frequency domain transform coefficients; and
the polarized frequency domain transform coefficients are inverse frequency domain transformed to obtain three components Rn, Gn, Bn.
2. The image enhancement method of claim 1, further comprising:
the three components R, G, B are normalized before the three components R, G, B of the color image are frequency domain transformed, respectively.
3. The image enhancement method of claim 1, wherein the frequency domain transform is a Contourlet (Contourlet) transform and the frequency domain transform coefficients are Contourlet transform coefficients.
4. The image enhancement method according to claim 3, wherein the step of setting the frequency domain transform coefficients comprises: the frequency domain transform coefficients are set by the following equation,
where δ represents a predetermined threshold, Coeff represents a Contourlet transform coefficient, Coeff δ represents Coeff after setting, sgn represents a sign-taking operation, max represents a maximum-value-taking operation, mean2 represents a mean-value-taking operation, std2 represents a standard deviation-taking operation, and σ represents an estimated noise standard deviation value of R, G, B, the value of which is obtained by the equationA calculation is performed in which Median denotes the Median operation, Coeff1 denotes the first layer coefficient of Coeff, and r has a value of 0.6745.
5. The image enhancement method of claim 4, wherein δ employs a Bayesian shrinkage (Bayesian shrinkage) threshold.
6. The image enhancement method of claim 5, wherein δ employs an adaptive Bayesian shrinkage (Bayesian shrinkage) threshold.
7. The image enhancement method of claim 6, wherein the step of quantizing the set frequency domain transform coefficients comprises: the first layer coefficient in the set Contourlet transform coefficients is quantized.
8. The image enhancement method of claim 7, wherein the step of quantizing the first layer coefficient in the set Contourlet transform coefficients comprises:
selecting a two-dimensional area for each point in each of a plurality of subgraphs constituted by coefficients of Coeff δ converted to the same direction in the two-dimensional first-layer coefficients by a predetermined matrix;
performing convolution processing on each selected two-dimensional area through a mean matrix to obtain a convolution value;
for each selected two-dimensional region, the value of the center point takes the maximum value in the selected two-dimensional region when the value of the center point of the selected two-dimensional region is greater than or equal to the convolution value, and the value of the center point takes the minimum value in the selected two-dimensional region when the value of the center point of the selected two-dimensional region is less than the convolution value.
9. The image enhancement method of claim 1, further comprising:
spatial domain image enhancement is performed by using the three components Rn, Gn, Bn.
10. The image enhancement method of claim 9, wherein the step of performing spatial domain image enhancement by using the three components Rn, Gn, Bn comprises:
convolution processing is respectively carried out on the three components Rn, Gn and Bn through a matrix in a pyramid form to obtain RnF, GnF and BnF, the three components RNS, GNS and BNS of the spatial domain image enhancement are obtained through the following equations,
where Rn (i, j), Gn (i, j), and Bn (i, j) respectively represent values at coordinates (i, j) in Rn, Gn, Bn, RnF (i, j), GnF (i, j), BnF (i, j) respectively represent values at coordinates (i, j) in RnF, GnF, BnF, RNS, GNS (i, j), BNS (i, j) respectively represent values at coordinates (i, j) in RNS, GNS, BNS, t represents an overall saturation adjustment factor, ζ represents a local contrast adjustment factor, and RNS (i, j), GNS (i, j), BNS (i, j) are limited to corresponding end point values of a predetermined range when RNS (i, j), GNS (i, j), BNS (i, j) exceed the predetermined range.
11. The image enhancement method according to one of claims 1 to 10, wherein the color image is obtained by a portable terminal.
12. An image enhancement apparatus comprising:
a frequency domain transform unit that frequency domain transforms the three components R, G, B of the color image, respectively, to obtain frequency domain transform coefficients;
a threshold processing unit that sets the frequency domain transform coefficient whose absolute value is smaller than a predetermined threshold to zero, and sets the frequency domain transform coefficient whose absolute value is greater than or equal to the predetermined threshold to a value related to itself in consideration of continuity of the color image weak detail signal and undistorted strong detail signal;
a maximum and minimum processing unit, which increases the difference by polarizing the set frequency domain transformation coefficient; and
and a frequency domain inverse transformation unit which performs frequency domain inverse transformation on the polarized frequency domain transformation coefficients to obtain three components Rn, Gn and Bn.
13. The image enhancement apparatus of claim 12, further comprising:
and a spatial domain image enhancement unit for performing spatial domain image enhancement by using the three components Rn, Gn, Bn.
14. The image intensifier as set forth in one of claims 12 and 13, wherein the color image is obtained by a portable terminal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310103337.0A CN104077744B (en) | 2013-03-27 | 2013-03-27 | Image enhancing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310103337.0A CN104077744B (en) | 2013-03-27 | 2013-03-27 | Image enhancing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104077744A true CN104077744A (en) | 2014-10-01 |
| CN104077744B CN104077744B (en) | 2017-02-08 |
Family
ID=51598987
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310103337.0A Expired - Fee Related CN104077744B (en) | 2013-03-27 | 2013-03-27 | Image enhancing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104077744B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105447830A (en) * | 2015-11-27 | 2016-03-30 | 合一网络技术(北京)有限公司 | Method and apparatus for strengthening clarity of dynamic video image |
| CN112446833A (en) * | 2019-09-02 | 2021-03-05 | 武汉Tcl集团工业研究院有限公司 | Image processing method, intelligent terminal and storage medium |
| CN113627232A (en) * | 2021-06-17 | 2021-11-09 | 浙江科技学院 | Water dispenser control system and control method based on analysis of human body water shortage state |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101777181A (en) * | 2010-01-15 | 2010-07-14 | 西安电子科技大学 | Ridgelet bi-frame system-based SAR image airfield runway extraction method |
-
2013
- 2013-03-27 CN CN201310103337.0A patent/CN104077744B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101777181A (en) * | 2010-01-15 | 2010-07-14 | 西安电子科技大学 | Ridgelet bi-frame system-based SAR image airfield runway extraction method |
Non-Patent Citations (4)
| Title |
|---|
| PENG FENG: "《Enhancing retinal image by the Contourlet transform》", 《PATTERN RECOGNITION LETTERS》 * |
| SHAO WEI DAI等: "《IMAGE DENOISING BASED ON COMPLEX CONTOURLET TRANSFORM》", 《PROCEEDINGS OF THE 2007 INTERNATIONAL CONFERENCE ON WAVELET ANALYSIS AND PATTERN RECOGNITION》 * |
| 金彩虹: "《基于非下采样Contourlet变换的图像自适应阈值去噪算法》", 《华中师范大学学报(自然科学版)》 * |
| 韩晓娜等: "《基于Contourlet变换的图像去噪算法》", 《航空工程进展》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105447830A (en) * | 2015-11-27 | 2016-03-30 | 合一网络技术(北京)有限公司 | Method and apparatus for strengthening clarity of dynamic video image |
| CN105447830B (en) * | 2015-11-27 | 2018-05-25 | 合一网络技术(北京)有限公司 | Dynamic video image clarity intensifying method and device |
| CN112446833A (en) * | 2019-09-02 | 2021-03-05 | 武汉Tcl集团工业研究院有限公司 | Image processing method, intelligent terminal and storage medium |
| CN112446833B (en) * | 2019-09-02 | 2024-05-31 | 武汉Tcl集团工业研究院有限公司 | Image processing method, intelligent terminal and storage medium |
| CN113627232A (en) * | 2021-06-17 | 2021-11-09 | 浙江科技学院 | Water dispenser control system and control method based on analysis of human body water shortage state |
| CN113627232B (en) * | 2021-06-17 | 2024-02-09 | 浙江科技学院 | Water dispenser control system and control method based on analysis of water shortage state of human body |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104077744B (en) | 2017-02-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6057881B2 (en) | Method for removing noise from input image consisting of pixels containing noise | |
| Demirel et al. | Satellite image contrast enhancement using discrete wavelet transform and singular value decomposition | |
| Hemalatha et al. | Wavelet transform based steganography technique to hide audio signals in image | |
| Cooper | Improved photo response non-uniformity (PRNU) based source camera identification | |
| Saxena et al. | Noises and image denoising techniques: A brief survey | |
| Kim et al. | Wavelet-domain color image enhancement using filtered directional bases and frequency-adaptive shrinkage | |
| Seiler et al. | Resampling images to a regular grid from a non-regular subset of pixel positions using frequency selective reconstruction | |
| Gupta et al. | Image denoising using wavelet transform method | |
| CN107784638A (en) | A kind of Dongba ancient books image enchancing method of optimization | |
| CN108596843B (en) | Underwater image color recovery algorithm based on bright channel | |
| KR20070085363A (en) | Degradation information restoring method and device | |
| Makwana et al. | Single image super-resolution via iterative back projection based Canny edge detection and a Gabor filter prior | |
| CN104077744B (en) | Image enhancing method and device | |
| CN105338219A (en) | Video image denoising processing method and apparatus | |
| CN111192204A (en) | Image enhancement method, system and computer readable storage medium | |
| CN109872289A (en) | Image enchancing method based on improved non-downsampling Contourlet conversion | |
| Akbarzadeh et al. | An efficient single image super resolution algorithm based on wavelet transforms | |
| Kim et al. | Vaguelette-wavelet decomposition for frequency adaptive image restoration using directional wavelet bases | |
| CN103839237B (en) | SAR image despeckling method based on SVD dictionary and linear minimum mean square error estimation | |
| CN106683055A (en) | Degradation model and group sparse representation-based foggy day image restoration method | |
| CN110517196B (en) | SAR image denoising method and system | |
| Kaur et al. | Performance evaluation of modified DBLA Using dark channel prior & CLAHE | |
| Sukhatme et al. | Independent component analysis based denoising of magnetic resonance images | |
| Guo et al. | An Image Denoising Algorithm based on Kuwahara Filter | |
| Kour et al. | Hybrid filter with wavelet denoising and anisotropic diffusion filter for image despeckling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170208 Termination date: 20210327 |