CN111770246A - Image noise reduction device and method - Google Patents
Image noise reduction device and method Download PDFInfo
- Publication number
- CN111770246A CN111770246A CN201910262684.5A CN201910262684A CN111770246A CN 111770246 A CN111770246 A CN 111770246A CN 201910262684 A CN201910262684 A CN 201910262684A CN 111770246 A CN111770246 A CN 111770246A
- Authority
- CN
- China
- Prior art keywords
- image
- noise reduction
- visible light
- corrected
- offset
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
- H04N23/81—Camera processing pipelines; Components thereof for suppressing or minimising disturbance in the image signal generation
- H04N23/811—Camera processing pipelines; Components thereof for suppressing or minimising disturbance in the image signal generation by dust removal, e.g. from surfaces of the image sensor or processing of the image signal output by the electronic image sensor
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an image noise reduction device and a method, wherein the device comprises: a data input unit for acquiring raw image data using an image sensor; the interpolation unit is used for interpolating the original image data to obtain four pieces of full-resolution data; the other component eliminating unit is used for executing the correction of the visible light color by subtracting other components in the R \ G \ B channel; the boundary retention noise reduction unit is used for fusing the overall brightness and color information of the corrected visible light image with the position information of the boundary details of the image which is not corrected, and utilizing the high signal-to-noise ratio characteristic of the image which is not corrected to achieve the noise reduction effect of boundary retention; and the output unit is used for outputting the RGB image subjected to denoising and fusion to a subsequent image processing module.
Description
Technical Field
The present invention relates to the field of image processing technologies, and in particular, to an image denoising device and method.
Background
At present, a common image sensor senses visible light and near infrared light simultaneously in order to improve low-light imaging effect. To eliminate the interference of infrared light with color, it is widely adopted in the industry to provide cameras with dual-filter switching devices (IR-CUTs) to correct daytime color.
RGBIR is a new filter arrangement for image sensors. Compared with the arrangement mode of a Bayer format filter commonly used by a traditional image sensor, RGBIR replaces part of green filters in the Bayer format with infrared filters (Infra Red, IR for short). The advantage of the rgbiir format image sensor over the Bayer format image sensor is that it can sense both visible and non-visible light, and fig. 1 is a common rgbiir image sensor format.
To eliminate the infrared interference on the color, the rgbiir sensor-based device eliminates the infrared interference through a correlation algorithm. The basic idea of the common de-infrared algorithm is to try to calculate the full-width infrared image and the RGB image affected by the infrared respectively, and then subtract the infrared image from the RGB channel respectively to obtain the processed visible light image. Although these algorithms can eliminate the infrared interference, they lose the high signal-to-noise ratio of the RGB images affected by infrared, especially in the case of low-light and low-color temperature, so that in industrial applications, there is often a balance between noise and color accuracy in the case of low-light or low-color temperature.
Disclosure of Invention
In order to overcome the above-mentioned deficiencies of the prior art, an object of the present invention is to provide an image noise reduction apparatus and method, which fuse the color and the overall brightness information of the corrected (e.g. infrared corrected) visible light image with the position information of the boundary details of the image that is not corrected (e.g. infrared corrected) to achieve the noise reduction effect of boundary preservation.
To achieve the above and other objects, the present invention provides an image noise reduction apparatus, comprising:
a data input unit for acquiring raw image data using an image sensor;
the interpolation unit is used for interpolating the original image data to obtain four pieces of full-resolution data;
the other component eliminating unit is used for executing the correction of the visible light color by subtracting other components in the R \ G \ B channel;
the boundary retention noise reduction unit is used for fusing the overall brightness and color information of the corrected visible light image with the position information of the boundary details of the image which is not corrected, and utilizing the high signal-to-noise ratio characteristic of the image which is not corrected to achieve the noise reduction effect of boundary retention;
and the output unit is used for outputting the RGB image subjected to denoising and fusion to a subsequent image processing module.
Preferably, the data input unit acquires original image data in an rgbiir format by using an rgbiir image sensor, and the interpolation unit interpolates the acquired data in the rgbiir format to obtain data with four resolutions of R \ G \ B \ IR.
Preferably, the other component removing unit separates infrared of RGB channels and visible light of infrared channels using an infrared reduction matrix for the entire image.
Preferably, the infrared reduction matrix adopts a 4 × 4 correction matrix as follows:
wherein, aijAre the correction parameters.
Preferably, the boundary-preserving noise reduction unit further includes:
the first format conversion module is used for converting the uncorrected image I and the corrected visible light image P to a luminance chrominance domain to respectively obtain a luminance graph and a UV chrominance graph of the two images;
the phase information extraction module is used for extracting phase information of a brightness map of the uncorrected image I;
the spectrum information extraction module is used for extracting the spectrum information of the brightness map of the corrected visible light image P;
the luminance map fusion module is used for fusing the phase information of the luminance map of the uncorrected image I and the frequency spectrum information of the luminance map of the corrected visible light image P so as to restore the image, and the restored image has the detail boundary of the uncorrected image I and the luminance of the corrected visible light image P;
the chromaticity diagram noise reduction module is used for carrying out noise reduction processing on the chromaticity diagram of the corrected visible light image P;
and the second format conversion module is used for fusing the brightness map obtained by fusion and the chroma map subjected to noise reduction processing and converting the fused brightness map and the chroma map into an RGB domain.
Preferably, the first format conversion module performs format conversion by using the following conversion formula:
Y＝0.30R+0.59G+0.11B
U＝0.493(B-Y)+Offset
V＝0.877(R-Y)+Offset
wherein Offset is an Offset.
Preferably, the phase information extraction module and the spectrum information extraction module extract the phase information and the spectrum information by performing two-dimensional fast fourier transform on the image.
Preferably, the luminance map fusion module is configured to perform inverse fast fourier transform on the phase information of the luminance map of the uncorrected image I and the frequency spectrum information of the luminance map of the corrected visible light image P to perform image restoration.
Preferably, the second format conversion module performs format conversion by using the following conversion formula:
R＝Y+1.4075(V-Offset)
G＝Y-0.3455(U-Offset)-0.7169(V-Offset)
B＝Y-1.779(U-Offset)
wherein Offset is an Offset.
In order to achieve the above object, the present invention further provides an image noise reduction method, including the following steps:
step S1, acquiring raw image data using an image sensor;
step S2, interpolating the original image data to obtain four full-resolution data;
step S3, correcting the color of the visible light by subtracting other components in the R \ G \ B channel;
step S4, fusing the whole brightness and color information of the corrected visible light image with the position information of the boundary details of the image which is not corrected, and utilizing the high signal-to-noise ratio characteristic of the image which is not corrected to achieve the noise reduction effect of boundary reservation;
and S5, outputting the RGB image subjected to denoising and fusion to a subsequent image processing module.
Compared with the prior art, the image noise reduction device and the method respectively extract phase information and spectrum information from an uncorrected (such as infrared correction) RGB image and a corrected (such as infrared correction) RGB image luminance image through fast Fourier transform, recover the image through fast Fourier inverse transform to obtain a denoised luminance image, and fuse the denoised luminance image and the corrected RGB image chrominance image.
Drawings
FIG. 1 is a schematic diagram of a format of an image sensor of RGBIR;
FIG. 2 is a system architecture diagram of an image noise reduction apparatus of the present invention;
fig. 3 is a detailed structure diagram of the boundary preserving noise reduction unit 104 according to the embodiment of the invention;
FIG. 4 is a flowchart illustrating steps of a method for image denoising according to the present invention;
FIG. 5 is a detailed flowchart of step S4 according to the embodiment of the present invention;
FIG. 6 is a flow chart of an embodiment of the present invention.
Detailed Description
Other advantages and capabilities of the present invention will be readily apparent to those skilled in the art from the present disclosure by describing the embodiments of the present invention with specific embodiments thereof in conjunction with the accompanying drawings. The invention is capable of other and different embodiments and its several details are capable of modification in various other respects, all without departing from the spirit and scope of the present invention.
Fig. 2 is a system architecture diagram of an image noise reduction apparatus according to the present invention. As shown in fig. 2, an image noise reduction device of the present invention includes:
a data input unit 101 for acquiring raw image data with an image sensor and outputting to subsequent units. In the embodiment of the present invention, the data input unit 101 employs an RGBIR image sensor to acquire an RGBIR image and process the RGBIR image into RGBIR format data to be transmitted to a subsequent unit, but it should be noted that the present invention may also acquire image data in other formats, such as RGBW format, RCCB format, etc., and the present invention is not limited thereto.
The interpolation unit 102 is configured to interpolate the obtained original image data to obtain four full-resolution data for a subsequent unit to perform color correction processing, in a specific embodiment of the present invention, the interpolation unit 102 interpolates the obtained rgbiir format data to obtain data with four resolutions of R \ G \ B \ IR, where the interpolation unit 102 may be a common interpolation method such as bilinear (bilinear) and bicubic (bicubic), or may be implemented by other advanced interpolation methods, and the present invention is not limited thereto.
And the other component eliminating unit 103 is used for executing the correction of the visible light color by subtracting other components in the R \ G \ B channel. Taking the obtained data of four resolutions of R \ G \ B \ IR as an example, specifically, the infrared and visible light of the infrared channel of the RGB channel are separated from the whole image by using an infrared reduction matrix, in a specific embodiment of the present invention, the infrared reduction matrix adopts a 4 × 4 correction matrix, as follows:
wherein the content of the first and second substances,
in order to obtain image data before infrared elimination,
for the image data after the infrared elimination,
the infrared-reducing matrix
Correction parameter a ofijThe calibration can be obtained by using a least square method according to the ideal spectral characteristics and the actual spectral characteristics; the color chart data shot by the standard 650nm filter and the 850nm filter under the same illumination condition can be obtained by using a least square method or other higher-level methods, and the invention is not limited to the method.
The boundary preserving and denoising unit 104 is configured to fuse the entire brightness and color information of the corrected visible light image with the position information of the boundary details of the image that is not corrected, and utilize the high snr characteristic of the image that is not corrected to achieve the noise reduction effect of boundary preserving, taking infrared correction as an example, that is, the boundary preserving and denoising unit 104 fuses the entire brightness and color information of the visible light image that is infrared corrected with the position information of the boundary details of the image that is not infrared corrected, specifically, as shown in fig. 3, the boundary preserving and denoising unit 104 further includes:
and a format conversion module 1041, configured to convert the uncorrected image I and the corrected visible light image P into a luminance and chrominance domain, so as to obtain a luminance graph and a UV chrominance graph of the two images, respectively. Still taking RGBIR as an example, the following expression can be used for the conversion formula:
Y＝0.30R+0.59G+0.11B
U＝0.493(B-Y)+Offset
V＝0.877(R-Y)+Offset
offset is an Offset added to facilitate integer arithmetic. Since the color of the image I without ir correction is interfered by ir, the chrominance information of the image I is useless, and if the luminance information Y of the image I is directly fused with the chrominance information UV of the visible light image P, the luminance information Y of the image I is not matched with the chrominance information UV of the visible light image P due to the different reflectivity of the objects in the scene to ir, for example, the luminance of the object with strong ir reflectivity is abnormally high, and the saturation thereof is obviously low.
The phase information extracting module 1042 is configured to extract phase information of a luminance map of the uncorrected image I. Still taking the RGBIR format as an example, in the specific embodiment of the present invention, the phase is represented by the following formula:
wherein, I (u, v) is an image imaginary part, and R (u, v) is a figure.
The spectral content of an image determines the amplitude of the sinusoids that form an image by combination, and the phase carries information about the displacement of the different sinusoids with respect to their respective origin, so that the phase information dominates the shape characteristics of the image when reconstructing the image. Because the two-dimensional fourier transform has separability, when the fourier transform is performed on the luminance information Y of the image I, the specific implementation can be as follows: the phase information extraction module 1042 takes a one-dimensional fast fourier transform along each row and then takes a one-dimensional fast fourier transform along each column. The fast fourier transform is based on discrete fourier transform, and uses the periodicity and symmetry of its coefficients to reduce the amount of computation, and can be basically divided into two categories: the time extraction method and the frequency extraction method are not specifically described here.
A spectrum information extracting module 1043, configured to extract spectrum information of the luminance map of the corrected visible light image P. In the embodiment of the invention, similarly, the two-dimensional fast fourier transform is performed on the image, and the frequency spectrum of the two-dimensional fast fourier transform is represented by the following formula:
and the luminance map fusion module 1044 is configured to perform inverse fast fourier transform on the phase information of the luminance map of the uncorrected image I and the frequency spectrum information of the luminance map of the corrected visible light image P, so as to perform image restoration. The restored image has both the detail boundaries of the uncorrected image I and the luminance of the corrected visible light image P.
And a chromaticity diagram noise reduction module 1045, configured to perform noise reduction processing on the chromaticity diagram of the corrected visible light image P. The color noise of the visible light image P corrected under low illumination is usually very large, and further filtering and noise reduction processing is required. Specifically, the denoising method can be a denoising means such as gaussian filtering or wavelet denoising, and under the low-light condition, the lower sampling and then the filtering operation can be performed on the chromaticity diagram, and finally the upper sampling is performed.
And a format conversion module 1046, configured to fuse the luminance map obtained by fusing with the chroma map subjected to noise reduction processing, and convert the fused luminance map into an RGB domain. Specifically, the conversion formula is determined by the above formula of RGB2YUV, and the following expression can be adopted:
R＝Y+1.4075(V-Offset)
G＝Y-0.3455(U-Offset)-0.7169(V-Offset)
B＝Y-1.779(U-Offset)。
wherein Offset is an Offset.
And the output unit 105 is configured to send the denoised and fused RGB image to a subsequent image processing module, and perform format rearrangement on the data through interpolation according to subsequent requirements.
It should be noted that the FFT is only one specific embodiment of the present invention, and other transform domain approaches are also within the scope of the present invention.
FIG. 4 is a flowchart illustrating steps of an image denoising method according to the present invention. As shown in fig. 4, the image denoising method of the present invention includes the following steps:
in step S1, raw image data is acquired by the image sensor. In a specific embodiment of the present invention, an image is captured by using an RGBIR image sensor and processed into data in RGBIR format to be transmitted to a subsequent unit.
In step S2, the original image data is interpolated to obtain four full-resolution data for the subsequent color correction processing. In the specific embodiment of the present invention, the original image data in the rgbiir format is taken as an example, the obtained data in the rgbiir format is interpolated to obtain data with four resolutions of R \ G \ B \ IR, and the interpolation may be implemented by a common interpolation method such as bilinear (bilinear) and bicubic (bicubic), or other advanced interpolation, which is not limited in this respect.
And step S3, correcting the color of the visible light by subtracting other components in the R \ G \ B channel. Specifically, taking the rgbiir format as an example, the infrared and visible light of the infrared channels of the RGB channels are separated from the whole image by using an infrared reduction matrix, in a specific embodiment of the present invention, the infrared reduction matrix adopts a 4 × 4 rectification matrix, as follows:
the parameter calibration of the infrared reduction matrix can be obtained by utilizing a least square method according to the ideal spectral characteristics and the actual spectral characteristics; the color chart data shot by the standard 650nm filter and the 850nm filter under the same illumination condition can be obtained by using a least square method or other higher-level methods, and the invention is not limited to the method.
And step S4, fusing the whole brightness and color information of the corrected visible light image with the position information of the boundary details of the image which is not corrected, and utilizing the high signal-to-noise ratio characteristic of the image which is not corrected to achieve the noise reduction effect of boundary retention. Taking the RGBIR format as an example, the overall brightness and color information of the visible light image after infrared correction is fused with the position information of the boundary details of the image without infrared correction, and the high signal-to-noise ratio characteristic of the image without infrared correction is utilized to achieve the noise reduction effect of boundary retention.
Specifically, as shown in fig. 5, step S4 further includes:
step S400, the uncorrected image I and the corrected visible light image P are transferred to a luminance chromaticity domain, and a luminance graph and a UV chromaticity graph of the two images are obtained respectively. Taking the rgbiir format as an example, the following expression can be adopted for the conversion formula:
Y＝0.30R+0.59G+0.11B
U＝0.493(B-Y)+Offset
V＝0.877(R-Y)+Offset
since the color of the image I without ir correction is interfered by ir, the chrominance information of the image I is useless, and if the luminance information Y of the image I is directly fused with the chrominance information UV of the visible light image P, the luminance information Y of the image I is not matched with the chrominance information UV of the visible light image P due to the different reflectivity of the objects in the scene to ir, for example, the luminance of the object with strong ir reflectivity is abnormally high, and the saturation thereof is obviously low.
In step S401, phase information of the luminance map of the uncorrected image I is extracted. In a specific embodiment of the present invention, the phase is represented by the following formula:
wherein, I (u, v) is an image imaginary part, and R (u, v) is a figure.
The spectral content of an image determines the amplitude of the sinusoids that form an image by combination, and the phase carries information about the displacement of the different sinusoids with respect to their respective origin, so that the phase information dominates the shape characteristics of the image when reconstructing the image. Because the two-dimensional fourier transform has separability, when the fourier transform is performed on the luminance information Y of the image I, the specific implementation can be as follows: one-dimensional fast fourier transforms are taken along each row and one-dimensional fast fourier transforms are taken along each column. The fast fourier transform is based on discrete fourier transform, and uses the periodicity and symmetry of its coefficients to reduce the amount of computation, and can be basically divided into two categories: the time extraction method and the frequency extraction method are not specifically described here.
In step S402, spectrum information of the luminance map of the corrected visible light image P is extracted. In the embodiment of the invention, similarly, the two-dimensional fast fourier transform is performed on the image, and the frequency spectrum of the two-dimensional fast fourier transform is represented by the following formula:
in step S403, inverse fast fourier transform is performed on the phase information of the luminance map of the uncorrected image I and the frequency spectrum information of the luminance map of the corrected visible light image P, so as to perform image restoration. The restored image has both the detail boundaries of the uncorrected image I and the luminance of the corrected visible light image P.
In step S404, noise reduction processing is performed on the chromaticity diagram of the corrected visible light image P. The color noise of the visible light image P corrected under low illumination is usually very large, and further filtering and noise reduction processing is required. Specifically, the denoising method can be a denoising means such as gaussian filtering or wavelet denoising, and under the low-light condition, the lower sampling and then the filtering operation can be performed on the chromaticity diagram, and finally the upper sampling is performed.
Step S405, the luminance graph obtained by fusion and the chroma graph subjected to noise reduction processing are fused and returned to the RGB domain. Specifically, the conversion formula is determined by the above formula of RGB2YUV, and may be the following expression:
R＝Y+1.4075(V-Offset)
G＝Y-0.3455(U-Offset)-0.7169(V-Offset)
B＝Y-1.779(U-Offset)。
and step S5, sending the RGB image subjected to denoising and fusion to a subsequent image processing module, and performing format rearrangement on data through interpolation according to subsequent requirements.
It should be noted that the FFT is only one specific embodiment of the present invention, and other transform domain approaches are also within the scope of the present invention.
FIG. 6 is a flow chart of an embodiment of the present invention. In the embodiment of the present invention, taking the color preserving image denoising method based on the rgbiir sensor as an example, the steps are as follows:
in step 201, an RGBIR image sensor is used to acquire image data to transmit data in an RGBIR format to a subsequent unit.
Step 203, infrared elimination: the correction of the visible light color is executed by subtracting the infrared component in the R \ G \ B channel, specifically, the visible light of the infrared channel and the infrared channel of the RGB channel are separated by using a 4 multiplied by 4 correction matrix for the whole image. The formula is as follows:
Step 207, fusing the brightness map: and performing inverse fast Fourier transform on the phase information of the brightness map of the image I without infrared correction and the frequency spectrum information of the brightness map of the visible light image P after infrared correction, and performing image restoration.
And step 210, outputting, namely sending the RGB images subjected to denoising and fusion to a subsequent image processing module, and performing format rearrangement on data through interpolation according to subsequent requirements.
In summary, the image noise reduction apparatus and method of the present invention extract phase information and spectral information from the uncorrected (e.g., infrared corrected) RGB image and the corrected (e.g., infrared corrected) RGB image by fast fourier transform, perform image restoration by inverse fast fourier transform to obtain the denoised luminance image, and fuse the denoised luminance image with the corrected RGB chrominance image.
The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Modifications and variations can be made to the above-described embodiments by those skilled in the art without departing from the spirit and scope of the present invention. Therefore, the scope of the invention should be determined from the following claims.
Claims (10)
1. An image noise reduction apparatus comprising:
a data input unit for acquiring raw image data using an image sensor;
the interpolation unit is used for interpolating the original image data to obtain four pieces of full-resolution data;
the other component eliminating unit is used for executing the correction of the visible light color by subtracting other components in the R \ G \ B channel;
the boundary retention noise reduction unit is used for fusing the overall brightness and color information of the corrected visible light image with the position information of the boundary details of the image which is not corrected, and utilizing the high signal-to-noise ratio characteristic of the image which is not corrected to achieve the noise reduction effect of boundary retention;
and the output unit is used for outputting the RGB image subjected to denoising and fusion to a subsequent image processing module.
2. The image noise reduction apparatus according to claim 1, wherein: the data input unit acquires original image data in an RGBIR format by using an RGBIR image sensor, and the interpolation unit interpolates the acquired data in the RGBIR format to acquire data with four resolutions of R \ G \ B \ IR.
3. The image noise reduction apparatus according to claim 2, wherein: the other component eliminating unit separates infrared of the RGB channel and visible light of the infrared channel from the whole image by using the infrared reduction matrix.
4. The image noise reduction apparatus according to claim 3, wherein: the infrared reduction matrix adopts the following correction matrix of 4 multiplied by 4:
wherein, aijAre the correction parameters.
5. The image noise reduction apparatus according to claim 1, wherein: the boundary preserving noise reduction unit further includes:
the first format conversion module is used for converting the uncorrected image I and the corrected visible light image P to a luminance chrominance domain to respectively obtain a luminance graph and a UV chrominance graph of the two images;
the phase information extraction module is used for extracting phase information of a brightness map of the uncorrected image I;
the spectrum information extraction module is used for extracting the spectrum information of the brightness map of the corrected visible light image P;
the luminance map fusion module is used for fusing the phase information of the luminance map of the uncorrected image I and the frequency spectrum information of the luminance map of the corrected visible light image P so as to restore the image, and the restored image has the detail boundary of the uncorrected image I and the luminance of the corrected visible light image P;
the chromaticity diagram noise reduction module is used for carrying out noise reduction processing on the chromaticity diagram of the corrected visible light image P;
and the second format conversion module is used for fusing the brightness map obtained by fusion and the chroma map subjected to noise reduction processing and converting the fused brightness map and the chroma map into an RGB domain.
6. The image noise reduction apparatus according to claim 5, wherein the first format conversion module performs format conversion using a conversion formula:
Y＝0.30R+0.59G+0.11B
U＝0.493(B-Y)+Offset
V＝0.877(R-Y)+Offset
wherein Offset is an Offset.
7. The image noise reduction apparatus according to claim 5, wherein the phase information extraction module and the spectrum information extraction module extract the phase information and the spectrum information by performing two-dimensional fast Fourier transform on the image.
8. The image noise reduction apparatus according to claim 5, wherein: the luminance map fusion module is used for performing inverse fast Fourier transform on the phase information of the luminance map of the uncorrected image I and the frequency spectrum information of the luminance map of the corrected visible light image P to perform image restoration.
9. The image noise reduction apparatus according to claim 5, wherein the second format conversion module performs format conversion using a conversion formula:
R＝Y+1.4075(V-Offset)
G＝Y-0.3455(U-Offset)-0.7169(V-Offset)
B＝Y-1.779(U-Offset)
wherein Offset is an Offset.
10. An image noise reduction method comprising the steps of:
step S1, acquiring raw image data using an image sensor;
step S2, interpolating the original image data to obtain four full-resolution data;
step S3, correcting the color of the visible light by subtracting other components in the R \ G \ B channel;
step S4, fusing the whole brightness and color information of the corrected visible light image with the position information of the boundary details of the image which is not corrected, and utilizing the high signal-to-noise ratio characteristic of the image which is not corrected to achieve the noise reduction effect of boundary reservation;
and step S5, outputting the RGB image subjected to denoising and fusion to a subsequent image processing module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910262684.5A CN111770246A (en) | 2019-04-02 | 2019-04-02 | Image noise reduction device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910262684.5A CN111770246A (en) | 2019-04-02 | 2019-04-02 | Image noise reduction device and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111770246A true CN111770246A (en) | 2020-10-13 |
Family
ID=72718524
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910262684.5A Pending CN111770246A (en) | 2019-04-02 | 2019-04-02 | Image noise reduction device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111770246A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130016905A1 (en) * | 2011-07-12 | 2013-01-17 | Park Byung-Kwan | Method and apparatus for correcting color distortion |
| CN103413127A (en) * | 2013-09-10 | 2013-11-27 | 上海海事大学 | Marine target significance detection method based on spectrum singular value decomposition |
| CN103686111A (en) * | 2013-12-31 | 2014-03-26 | 上海富瀚微电子有限公司 | Method and device for correcting color based on RGBIR (red, green and blue, infra red) image sensor |
| US20150296146A1 (en) * | 2011-06-10 | 2015-10-15 | Flir Systems, Inc. | Electrical cabinet infrared monitor systems and methods |
| CN106488201A (en) * | 2015-08-28 | 2017-03-08 | 杭州海康威视数字技术股份有限公司 | A kind of processing method of picture signal and system |
| CN106952245A (en) * | 2017-03-07 | 2017-07-14 | 深圳职业技术学院 | A kind of processing method and system for visible images of taking photo by plane |
| CN107967668A (en) * | 2016-10-20 | 2018-04-27 | 上海富瀚微电子股份有限公司 | A kind of image processing method and device |
-
2019
- 2019-04-02 CN CN201910262684.5A patent/CN111770246A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150296146A1 (en) * | 2011-06-10 | 2015-10-15 | Flir Systems, Inc. | Electrical cabinet infrared monitor systems and methods |
| US20130016905A1 (en) * | 2011-07-12 | 2013-01-17 | Park Byung-Kwan | Method and apparatus for correcting color distortion |
| CN103413127A (en) * | 2013-09-10 | 2013-11-27 | 上海海事大学 | Marine target significance detection method based on spectrum singular value decomposition |
| CN103686111A (en) * | 2013-12-31 | 2014-03-26 | 上海富瀚微电子有限公司 | Method and device for correcting color based on RGBIR (red, green and blue, infra red) image sensor |
| CN106488201A (en) * | 2015-08-28 | 2017-03-08 | 杭州海康威视数字技术股份有限公司 | A kind of processing method of picture signal and system |
| CN107967668A (en) * | 2016-10-20 | 2018-04-27 | 上海富瀚微电子股份有限公司 | A kind of image processing method and device |
| CN106952245A (en) * | 2017-03-07 | 2017-07-14 | 深圳职业技术学院 | A kind of processing method and system for visible images of taking photo by plane |
Non-Patent Citations (3)
| Title |
|---|
| 王金武等: "基于相位谱和频率调谐的海上场景显著性检测", 《计算机应用》 * |
| 章学静: "像素级图像增强及配准算法研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 肖钟捷等: "基于显著边缘信息的大气湍流图像盲复原", 《光电工程》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101531709B1 (en) | Image processing apparatus for generating high sensitive color image and method thereof | |
| CN107967668B (en) | Image processing method and device | |
| US8457433B2 (en) | Methods and systems for image noise filtering | |
| EP1580982A2 (en) | Image data processing in color spaces | |
| US20070183681A1 (en) | Adaptive image filter for filtering image information | |
| JP2023025085A (en) | Camera image processing method and camera | |
| JP2012060641A (en) | Digital raw image demosaic method, computer program thereof, and image sensor circuit or graphic circuit thereof | |
| CN111510691B (en) | Color interpolation method and device, equipment and storage medium | |
| GB2345217A (en) | Colour video image sensor | |
| US11423514B2 (en) | Image processing noise reduction | |
| US20030117507A1 (en) | Color filter array interpolation | |
| US8471933B2 (en) | Image processing apparatus, image processing method and computer program | |
| KR101923957B1 (en) | Image Processing Apparatus and Method for Improving Sensitivity | |
| US20150206280A1 (en) | Image processing apparatus, image processing method, and program | |
| EP3275169B1 (en) | Downscaling a digital raw image frame | |
| US20090009831A1 (en) | Image Processing Method, Image Capture Device, and Computer Readable Medium | |
| WO2015133130A1 (en) | Video capturing device, signal separation device, and video capturing method | |
| US20140037207A1 (en) | System and a method of adaptively suppressing false-color artifacts | |
| JP5259616B2 (en) | Image processing system and image processing program | |
| US20140118580A1 (en) | Image processing device, image processing method, and program | |
| CN111770246A (en) | Image noise reduction device and method | |
| US9071803B2 (en) | Image processing apparatus, image pickup apparatus, image processing method and non-transitory computer-readable storage medium storing image processing program | |
| CN115471420A (en) | Image processing device, imaging apparatus, method, electronic apparatus, and storage medium | |
| Park | Architectural analysis of a baseline isp pipeline | |
| Shi et al. | Region-adaptive demosaicking with weighted values of multidirectional information |
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 |