WO2012129897A1 - 彩色图像去噪声的方法及其系统 - Google Patents
彩色图像去噪声的方法及其系统 Download PDFInfo
- Publication number
- WO2012129897A1 WO2012129897A1 PCT/CN2011/080040 CN2011080040W WO2012129897A1 WO 2012129897 A1 WO2012129897 A1 WO 2012129897A1 CN 2011080040 W CN2011080040 W CN 2011080040W WO 2012129897 A1 WO2012129897 A1 WO 2012129897A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- channel
- image
- information
- denoising
- cross
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000013501 data transformation Methods 0.000 claims abstract description 9
- 238000007781 pre-processing Methods 0.000 claims description 22
- 230000001629 suppression Effects 0.000 claims description 15
- 238000000605 extraction Methods 0.000 claims description 13
- 230000002146 bilateral effect Effects 0.000 claims description 6
- 238000000354 decomposition reaction Methods 0.000 claims description 6
- 238000009499 grossing Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 4
- 241000023320 Luma <angiosperm> Species 0.000 claims description 3
- OSWPMRLSEDHDFF-UHFFFAOYSA-N methyl salicylate Chemical compound COC(=O)C1=CC=CC=C1O OSWPMRLSEDHDFF-UHFFFAOYSA-N 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 9
- 230000004927 fusion Effects 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 6
- 230000006872 improvement Effects 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000000295 complement effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 229910044991 metal oxide Inorganic materials 0.000 description 2
- 150000004706 metal oxides Chemical class 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013144 data compression Methods 0.000 description 1
- 229910052745 lead Inorganic materials 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/10—Image enhancement or restoration using non-spatial domain filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
Definitions
- the present invention relates to the field of video processing, and more particularly to color image denoising techniques.
- Video image denoising (noise reduction) is an important part of video processing. Since the image is inevitably affected by the interference of the transmission medium, the external and ambient optical signals, electrical signals, and mechanical damage during the recording, data compression, storage process, and transmission process, the video is made when the video image reaches the display terminal.
- the image content (the luminance value and the color value of each pixel) carried by the signal changes, and these changes are called image noise because of their spatial and temporal randomness.
- CMOS Complementary Metal-Oxide Semiconductor
- CCD Charge Coupled Device
- an embodiment of the present invention provides a method for denoising a color image, comprising the following steps:
- At least one chroma channel image is denoised according to the cross channel invariant information.
- Embodiments of the present invention also provide a color image denoising system, including:
- An image decomposition unit for decomposing a color image into a luminance channel image and N chrominance channel images, N ⁇ 2;
- a cross-channel invariant information extracting unit configured to extract cross-channel invariant information from the luminance channel image decomposed by the image decomposition unit;
- the denoising unit is configured to perform denoising processing on the at least one chroma channel image decomposed by the image decomposing unit according to the cross channel invariant information obtained by the cross channel invariant information extracting unit.
- the cross-channel invariant information is extracted from the luminance channel image for denoising processing of the chroma channel image, which can eliminate false color edges or textures, reduce blur at the edges, and obtain better overall denoising effect.
- the brightness channel image has information such as edge and texture. When using this information to denoise the chroma channel image, it can prevent the denoising strength from being different from other chrominances because the edge or texture in a certain chrominance channel image is not obvious. Causes false colored edges or textures, or blurry edges.
- N chroma channel images are denoised by using the channel invariant information, and the best denoising effect can be obtained.
- Gaussian smoothing is used to implement local data preprocessing, which requires less computational resources and is faster.
- the local data pre-processing is implemented by using a bilateral filter, and a better edge-preserving performance can be obtained, and the processing effect is better.
- FIG. 1 is a schematic flow chart of a method for denoising a color image according to a first embodiment of the present invention
- FIG. 2 is a schematic flow chart of extracting cross-channel invariant information from a luminance channel image in a method for denoising a color image according to a second embodiment of the present invention
- FIG. 3 is a schematic flow chart of extracting cross-channel invariant information from a luminance channel image in a method for denoising a color image according to a third embodiment of the present invention
- FIG. 4 is a schematic structural diagram of a system for color image denoising according to a fourth embodiment of the present invention.
- FIG. 5 is a schematic diagram showing an internal structure of a cross-channel invariant information extracting unit in a fifth embodiment of the present invention.
- FIG. 6 is a schematic diagram showing an internal structure of a noise removing unit in a fifth embodiment of the present invention.
- Fig. 7 is a schematic view showing an internal structure of a noise removing unit in a sixth embodiment of the present invention.
- a first embodiment of the invention relates to a method of denoising a color image.
- 1 is a flow chart showing a method of denoising the color image. The method includes the following steps:
- step 101 the color image is decomposed into a luminance channel image and N chrominance channel images, N ⁇ 2.
- the chrominance image is a single channel image of image data or a multi-channel joint image, such as a UV component in a YUV color space, or a Cb and Cr component in a YCbCr color space, or a YPbPr color space. Pb and Pr components, and so on.
- cross-channel invariant information is extracted from the luma channel image.
- the cross-channel invariant information can be any local image features such as edge information, texture information, noise strength, and the like. It can be understood that the cross-channel invariant information is not limited to edge information, texture information, noise intensity, but also image information shared between other channels (which is represented by edge information, texture information), or obtained by simple operation and stable. The information of each channel (which is represented by the noise intensity), and so on.
- step 103 the at least one chroma channel image is subjected to denoising processing according to the cross channel invariant information.
- the N chroma channel images obtained in step 101 are denoised according to the cross-channel invariant information to obtain an optimal denoising effect.
- the texture has a chromaticity bias
- the partial chrominance channel can be denoised only by using the cross-channel invariant information obtained from the luminance channel image.
- the cross-channel invariant information is extracted from the luminance channel image for denoising processing of the chroma channel image, which can eliminate false color edges or textures, reduce blur at the edges, and obtain better overall denoising effect.
- the brightness channel image has information such as edge and texture.
- this information can prevent the denoising strength from being different from other chrominances because the edge or texture in a certain chrominance channel image is not obvious. Causes false colored edges or textures, or blurry edges.
- a second embodiment of the invention relates to a method of denoising a color image.
- the second embodiment is improved on the basis of the first embodiment, and the main improvement is that an efficient cross-channel invariant information extraction method including preprocessing is employed in step 102.
- step 102 further includes the sub-steps shown in FIG. 2.
- step 201 local data pre-processing is performed on the partial image of the luminance channel image, which is used to suppress the influence of partial noise.
- the local data pre-processing may be Gaussian smoothing or filtering with a bilateral filter or the like. Using Gaussian smoothing to achieve local data preprocessing requires less computational resources and faster speeds.
- the use of bilateral filters for local data preprocessing can achieve better edge-preserving performance and better processing results. It can be understood that other types of filters can also be used for local data preprocessing according to specific application scenarios.
- step 202 local image feature information collection is performed.
- Local feature information collection includes collecting local texture intensity information, texture direction information, and the like.
- the edge texture extraction operator can be used to collect local feature information of the image, including but not limited to soble operator, laplace operator, canny operator and Gabor filter, etc. Depending on the application.
- the soble operator is used to collect local feature information of the image, and the absolute value of the response Sh of the soble operator in the horizontal direction of the local data region is taken as the extracted horizontal direction information Fh, and the soble operator is placed in The absolute value of the response Sv in the vertical direction of the local data region is taken as the extracted vertical direction information Fv.
- the information representing the local feature is calculated as the cross-channel invariant information.
- step 103 further includes the following sub-steps:
- the local block mean V0, the center line mean V1, and the center column mean V2 of the chroma channel image are calculated.
- Q can have many forms, which can be set according to the actual application environment.
- Q(Fh, Fv) k - Max(Fh, Fv). Where k is the factor used to control the degree of invariant information fusion across channels.
- a third embodiment of the present invention relates to a method of denoising a color image.
- the third embodiment is improved on the basis of the first embodiment.
- the main improvement is that after the local data transformation is performed on the luminance channel image, the cross-channel invariant information is extracted in the transform domain.
- This technical solution can utilize the different characteristics of the signal in the transform domain to extract the feature information such as the edge texture of the partial image data more effectively and accurately, so as to better utilize the information to denoise the chroma channel image. Specifically:
- Step 102 further includes the sub-steps shown in Figure 3:
- local data transformation is performed on the luminance channel image.
- Local data transformation is intended to transform the representation of image data, such as Fast Fourier Transform (Fast Fourier) Transform, referred to as "FFT", wavelet transform, and two-dimensional discrete cosine transform (Discrete Cosine) Transform, referred to as "DCT" for short.
- FFT Fast Fourier Transform
- DCT discrete cosine Transform
- the cross-channel invariant information extraction of the transform domain is to extract the edge texture characteristics of the local image data more efficiently and accurately by using the different characteristics of the signal in the transform domain.
- step 302 noise interference suppression is performed in the transform domain obtained by the local data transform.
- This step is mainly for performing noise suppression operations on coefficients in the local transform domain, with the aim of improving the reliability of local feature information extraction.
- the noise suppression method in the transform domain is a relatively mature technology and will not be described here.
- Feature information extraction is mainly the mining of the relationship between energy coefficients and image textures in the transform domain.
- the two-dimensional DCT transform is performed on the luminance channel image
- the horizontal direction edge information of the image can be extracted from the row of the coefficient matrix obtained after the transformation, such as the sum of the absolute values of the first row coefficients of the two-dimensional DCT coefficient matrix.
- Fv the vertical direction edge information of the image can be extracted from the column of the coefficient matrix obtained after the transformation, for example, the sum of the absolute values of the first column coefficient of the coefficient matrix is Fh
- the image of the coefficient matrix obtained after the transformation can extract the image
- the texture information such as the number of non-zero coefficients in the coefficient matrix, is F0.
- step 103 The local chrominance image denoising of step 103 can be performed in the same manner as step 103 of the second embodiment.
- the method embodiments of the present invention can all be implemented in software, hardware, firmware, and the like. Regardless of whether the invention is implemented in software, hardware, or firmware, the instruction code can be stored in any type of computer-accessible memory (eg, permanent or modifiable, volatile or non-volatile, solid state Or non-solid, fixed or replaceable media, etc.).
- the instruction code can be stored in any type of computer-accessible memory (eg, permanent or modifiable, volatile or non-volatile, solid state Or non-solid, fixed or replaceable media, etc.).
- the memory can be, for example, a programmable array logic (Programmable) Array Logic ("PAL” for short), Random Access Memory (“RAM”) Programmable Read Only Memory (“PROM”), read-only memory (Read-Only) Memory, referred to as "ROM”), electrically erasable programmable read-only memory (Electrically Erasable Programmable) ROM, referred to as "EEPROM”), magnetic disk, optical disk, Digital Versatile Disc (“DVD”) and so on.
- PAL programmable array logic
- RAM Random Access Memory
- PROM Programmable Read Only Memory
- ROM Read-Only Memory
- EEPROM electrically erasable programmable read-only memory
- magnetic disk magnetic disk
- optical disk optical disk
- DVD Digital Versatile Disc
- a fourth embodiment of the invention relates to a system for color image denoising.
- 4 is a schematic structural view of a system for denoising the color image.
- the color image denoising system includes:
- the cross-channel invariant information extracting unit is configured to extract cross-channel invariant information from the luminance channel image decomposed by the image decomposition unit.
- the cross-channel invariant information includes, but is not limited to, image information shared by the respective channels (such as edge texture information), and one or the other information of each channel information (such as noise intensity) that can be obtained by simple operation and stabilized. Several.
- the denoising unit is configured to perform denoising processing on the at least one chroma channel image decomposed by the image decomposing unit according to the cross channel invariant information obtained by the cross channel invariant information extracting unit.
- the denoising unit performs denoising processing on the N chroma channel images based on the cross channel invariant information.
- the color image is divided into two types of channel images, a luminance image and a chrominance image, and the cross-channel invariant information extracted from the luminance image is utilized in the chrominance image denoising process, and the color is guided.
- the degree image dynamically uses different denoising strategies to suppress noise.
- the first embodiment is a method embodiment corresponding to the present embodiment, and the present embodiment can be implemented in cooperation with the first embodiment.
- the related technical details mentioned in the first embodiment are still effective in the present embodiment, and are not described herein again in order to reduce repetition. Accordingly, the related art details mentioned in the present embodiment can also be applied to the first embodiment.
- a fifth embodiment of the present invention relates to a system for color image denoising.
- the fifth embodiment is improved on the basis of the fourth embodiment, and the main improvement is that an efficient cross-channel invariant information extraction method including preprocessing is employed in the cross-channel invariant information extracting unit.
- the cross-channel invariant information extracting unit includes the following modules:
- the pre-processing module is configured to perform local data pre-processing on the partial image of the luminance channel image, and the local data pre-processing is used to suppress the influence of partial noise.
- Local data of luminance image is preprocessed by local data to suppress the influence of partial noise and improve the accuracy of local feature information collection.
- the complexity of this processing depends on the accuracy requirements of feature information collection and extraction, such as low precision requirements, simple The Gaussian smoothing filter can be used. If the accuracy requirement is relatively high, it is not unreasonable to use a better-effect filter to implement pre-processing, such as bilateral filters.
- the collecting module is configured to collect local image feature information on the processing result of the preprocessing module.
- the collection module can collect local image feature information using a soble operator, a laplace operator, a canny operator, a Gabor filter, and the like.
- an extraction module configured to calculate, according to the local image feature information output by the collection module, information representing the local feature as cross-channel invariant information.
- the denoising unit may include the module shown in Figure 6:
- a local block mean calculation module for calculating a local block mean value V0 in the chroma channel image.
- the local block center line mean calculation module is used to calculate the local block center line mean value V1 in the chroma channel image.
- the local block center column mean value calculation module is used to calculate the local block center column mean value V2 in the chroma channel image.
- the cross-channel invariant information fusion denoising module is used for fusing the local data information of the chroma image and the invariant information between the channels, and comprehensively completing the denoising of the chroma image.
- the local block mean calculation module, the local block center line mean calculation module, and the partial block center column mean calculation module may also be disposed not in the denoising unit but in other units, and the change is only a conventional means.
- the second embodiment is a method embodiment corresponding to the present embodiment, and the present embodiment can be implemented in cooperation with the second embodiment.
- the related technical details mentioned in the second embodiment are still effective in the present embodiment, and are not described herein again in order to reduce repetition. Accordingly, the related art details mentioned in the present embodiment can also be applied to the second embodiment.
- a sixth embodiment of the present invention relates to a system for color image denoising.
- the sixth embodiment is improved on the basis of the fourth embodiment.
- the main improvement is that after the local data transformation is performed on the luminance channel image, the cross-channel invariant information is extracted in the transform domain.
- This technical solution can utilize the different characteristics of the signal in the transform domain to extract the feature information such as the edge texture of the partial image data more effectively and accurately, so as to better utilize the information to denoise the chroma channel image.
- the cross-channel invariant information extracting unit includes the following modules:
- a transform module for performing local data transformation on the luminance channel image.
- the noise interference suppression module is configured to perform noise interference suppression on the processing result of the transformation module.
- the noise interference suppression module is mainly for performing noise suppression operations on coefficients in the local transform domain, with the aim of improving the reliability of local feature information extraction.
- the noise suppression method in the transform domain is a relatively mature technology and will not be described here.
- an extraction module configured to extract feature information from the processing result of the noise interference suppression module as cross-channel invariant information.
- the third embodiment is a method embodiment corresponding to the present embodiment, and the present embodiment can be implemented in cooperation with the third embodiment.
- the related technical details mentioned in the third embodiment are still effective in the present embodiment, and are not described herein again in order to reduce repetition. Accordingly, the related art details mentioned in the present embodiment can also be applied to the third embodiment.
- cross-channel invariant information and some specific denoising methods are mentioned, which are only for ease of understanding and implementation, and are not necessarily limited to these specific forms and denoising modes, as long as cross-channels are utilized.
- the invariant information combined with the fusion and denoising of the local data information of the chroma image can achieve certain effects, and belongs to the scope of protection of the present invention.
- each unit or module mentioned in each system embodiment of the present invention is a logical unit or module.
- a logical unit or module may be a physical unit or module, or may be a A physical unit or a part of a module may also be implemented in a plurality of physical units or a combination of modules.
- the physical implementation of the units or modules themselves is not the most important, and the functions of the units or modules of the logic are implemented. Combination is the key to solving the technical problems raised by the present invention.
- the above various system embodiments of the present invention do not introduce a unit or module that is not closely related to solving the technical problem proposed by the present invention, which does not indicate that the above system implementation does not exist. Other units or modules.
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
- Facsimile Image Signal Circuits (AREA)
Abstract
本发明涉及视频处理领域，公开了一种彩色图像去噪声的方法及其系统。本发明中，从亮度通道图像中提取跨通道不变信息，用于色度通道图像的去噪声处理，可以消除虚假彩色边缘或者纹理，减少边缘处的模糊，获得更好的整体去噪效果。跨通道不变信息可以是边缘信息、纹理信息、噪声强度等等。可以先对亮度通道图像进行局部数据变换，在变换域中提取跨通道不变信息，利用信号在变换域中表现出的不同特性，更有效地提取局部图像数据的边缘纹理等特征信息，从而更好地利用这些信息对色度通道图像进行去噪声处理。
Description
本发明涉及视频处理领域，特别涉及彩色图像去噪声技术。
视频图像去噪（降噪）是视频处理领域的一个重要内容。由于图像在摄录、数据压缩、存储过程、传输过程中将不可避免地受到传输介质、外界和环境的光信号、电信号、机械损伤的干扰影响，在视频图像到达显示终端的时候，使得视频信号所承载的图像内容（各个像素点的亮度数值和色彩数值）发生变化，这些变化因具有空间和时间上的随机性，因此被称为图像噪声。
传统的彩色图像去噪方法通常认为彩色图像是由三幅独立的灰度图像组合而成的。因此，传统的彩色图像去噪方法一般采用同一种去噪方法分别对三个通道进行处理。例如美国专利US20090251570A1就公开了一种在亮度和色度图像上分别进行去噪处理的图像去噪技术。
然而，传统的彩色图像去噪方法的去噪结果常常难以令人满意，存在残留彩色噪点，虚假彩色边缘或纹理、边缘模糊等等缺点。本发明的发明人认为其原因在于，彩色图像一般由互补型金属氧化物半导体（Complementary
Metal-Oxide Semiconductor，简称“CMOS”）或电荷藕合器件（Charge Coupled
Device，简称“CCD”）组成的传感器采集，彩色图像中RGB或者YUV三通道间既存在着极大的相关性，同时也表现出不同的噪声特性，传统的彩色图像去噪方法对真实彩色图像的这些特征缺乏考虑。
本发明的目的在于提供一种彩色图像去噪声的方法及其系统，可以消除虚假彩色边缘或者纹理，减少边缘处的模糊，获得更好的整体去噪效果。
为解决上述技术问题，本发明的实施方式提供了一种彩色图像去噪声的方法，包括以下步骤：
将彩色图像分解为一个亮度通道图像和N个色度通道图像，N≥2；
从亮度通道图像提取跨通道不变信息；
根据跨通道不变信息，对至少一个色度通道图像进行去噪声处理。
本发明的实施方式还提供了一种彩色图像去噪声的系统，包括：
图像分解单元，用于将彩色图像分解为一个亮度通道图像和N个色度通道图像，N≥2；
跨通道不变信息提取单元，用于从图像分解单元分解出的亮度通道图像中提取跨通道不变信息；
去噪声单元，用于根据跨通道不变信息提取单元所得的跨通道不变信息，对图像分解单元分解出的至少一个色度通道图像进行去噪声处理。
本发明实施方式与现有技术相比，主要区别及其效果在于：
从亮度通道图像中提取跨通道不变信息，用于色度通道图像的去噪声处理，可以消除虚假彩色边缘或者纹理，减少边缘处的模糊，获得更好的整体去噪效果。亮度通道图像中有边缘、纹理等信息，利用这些信息进行色度通道图像的去噪时，可以防止因为某个色度通道图像中边缘或纹理不明显而去噪力度与其它色度不同，最终导致虚假彩色边缘或者纹理，或边缘模糊化。
进一步地，对所有的N个色度通道图像都利用跨通道不变信息进行去噪声处理，可以得到最佳的去噪效果。
可选地，使用高斯平滑实现局部数据预处理，所需的计算资源较少，速度较快。
可选地，使用双边滤波器实现局部数据预处理，可以获得较好的保持边缘性能，处理的效果较好。
进一步地，对亮度通道图像进行局部数据变换后，在变换域中提取跨通道不变信息，可以利用信号在变换域中表现出的不同特性，更有效、准确地提取局部图像数据的边缘纹理等特征信息，从而更好地利用这些信息对色度通道图像进行去噪声处理。
图1是本发明第一实施方式中一种彩色图像去噪声的方法的流程示意图；
图2是本发明第二实施方式中一种彩色图像去噪声的方法中从亮度通道图像提取跨通道不变信息的流程示意图；
图3是本发明第三实施方式中一种彩色图像去噪声的方法中从亮度通道图像提取跨通道不变信息的流程示意图；
图4是本发明第四实施方式中一种彩色图像去噪声的系统的结构示意图；
图5是本发明第五实施方式中跨通道不变信息提取单元的一种内部结构示意图；
图6是本发明第五实施方式中去噪声单元的一种内部结构示意图；
图7是本发明第六实施方式中去噪声单元的一种内部结构示意图。
在以下的叙述中，为了使读者更好地理解本申请而提出了许多技术细节。但是，本领域的普通技术人员可以理解，即使没有这些技术细节和基于以下各实施方式的种种变化和修改，也可以实现本申请各权利要求所要求保护的技术方案。
为使本发明的目的、技术方案和优点更加清楚，下面将结合附图对本发明的实施方式作进一步地详细描述。
本发明第一实施方式涉及一种彩色图像去噪声的方法。图1是该彩色图像去噪声的方法的流程示意图。该方法包括以下步骤：
在步骤101中，将彩色图像分解为一个亮度通道图像和N个色度通道图像，N≥2。
在本发明的各实施方式中，色度图像是图像数据的单通道图像或者多通道联合的图像，比如YUV色彩空间中的UV分量，或者YCbCr色彩空间中的Cb和Cr分量，或者YPbPr色彩空间中的Pb和Pr分量，等等。
此后进入步骤102，从亮度通道图像提取跨通道不变信息。跨通道不变信息可以是任何局部图像特征，如边缘信息、纹理信息、噪声强度等等。可以理解，跨通道不变信息并不限于边缘信息、纹理信息、噪声强度，也可以是其它各个通道间共有的图像信息（其代表是边缘信息、纹理信息），或通过简单运算得到的并且稳定的各通道的信息（其代表是噪声强度），等等。
此后进入步骤103，根据跨通道不变信息，对至少一个色度通道图像进行去噪声处理。
在本发明的一个较佳实施方式中，根据跨通道不变信息，对步骤101所得的N个色度通道图像均进行去噪声处理，以便得到最佳的去噪效果。当然，可以理解，在某些应用场景下，例如纹理有色度偏向，为了提高效率，则可以只对部分色度通道利用从亮度通道图像得到的跨通道不变信息进行去噪声处理。
从亮度通道图像中提取跨通道不变信息，用于色度通道图像的去噪声处理，可以消除虚假彩色边缘或者纹理，减少边缘处的模糊，获得更好的整体去噪效果。
亮度通道图像中有边缘、纹理等信息，利用这些信息进行色度通道图像的去噪时，可以防止因为某个色度通道图像中边缘或纹理不明显而去噪力度与其它色度不同，最终导致虚假彩色边缘或者纹理，或边缘模糊化。
本发明第二实施方式涉及一种彩色图像去噪声的方法。
第二实施方式在第一实施方式的基础上进行了改进，主要改进之处在于在步骤102中采用了包括预处理在内的高效跨通道不变信息提取方法。
本实施方式中，步骤102进一步包括如图2所示的子步骤。
在步骤201中，对亮度通道图像的局部图像进行局部数据预处理，该局部数据预处理用于抑制部分噪声的影响。
局部数据预处理可以是高斯平滑或以双边滤波器滤波等等。使用高斯平滑实现局部数据预处理，所需的计算资源较少，速度较快。使用双边滤波器实现局部数据预处理，可以获得较好的保持边缘性能，处理的效果较好。可以理解，也可以根据具体的应用场景，使用其它类型的滤波器进行局部数据预处理。
此后进入步骤202，进行局部图像特征信息收集。
局部特征信息收集包括收集局部纹理强度信息和纹理方向信息等。在空域图像数据处理中，可以采用边缘纹理提取算子收集图像局部特征信息，这些算子包括但不限于soble算子、laplace算子、canny算子和Gabor滤波器等等，在具体实现中可以视应用选用。
在本发明的一个优选实例中，采用soble算子收集图像局部特征信息，将soble算子在该局部数据区域水平方向的响应Sh的绝对值作为提取出来的水平方向信息Fh，将soble算子在该局部数据区域垂直方向的响应Sv的绝对值作为提取出来的垂直方向信息Fv。
此后进入步骤203，根据局部图像特征信息收集的结果，计算表征局部特征的信息作为跨通道不变信息。
本实施方式中，步骤103进一步包拖括以下子步骤：
计算色度通道图像的局部块均值V0、中心行均值V1和中心列均值V2。
计算R = Fv*V1 + Fh*V2 +
Q(Fh，Fv)*V0，其中，Fv为水平方向信息，Fh为垂直方向信息，Fv和Fh是从亮度通道图像提取的跨通道不变信息，Q是Fh和Fv的函数。
此外，可以理解，Q可以有很多形式，具体可以根据实际应用环境来设定，在本发明一个比较简单的例子中，Q(Fh，Fv) = k -
max(Fh，Fv)。其中k是用于控制跨通道不变信息融合程度的因子。
融合的方法有很多，只要利用到了跨通道不变信息对色度通道图像进行噪声处理就可以，本发明实施方式中的例子仅仅是为了说明方便而举的一个简单例子。
本发明第三实施方式涉及一种彩色图像去噪声的方法。
第三实施方式在第一实施方式的基础上进行了改进，主要改进之处在于：对亮度通道图像进行局部数据变换后，在变换域中提取跨通道不变信息。这种技术方案可以利用信号在变换域中表现出的不同特性，更有效准确地提取局部图像数据的边缘纹理等特征信息，从而更好地利用这些信息对色度通道图像进行去噪声处理。具体地说：
步骤102进一步包括如图3所示的子步骤：
在步骤301中，对亮度通道图像进行局部数据变换。局部数据变换旨在转换图像数据的表示方式，比如快速傅里叶变换（Fast Fourier
Transform，简称“FFT”）、wavelet变换以及二维离散余弦变换（Discrete Cosine
Transform，简称“DCT”）等。变换域跨通道不变信息提取就是利用信号在变换域中表现出的不同特性，更有效准确地提取局部图像数据的边缘纹理特性。
此后进入步骤302，在局部数据变换所得的变换域中进行噪声干扰抑制。该步骤主要是为了对局部变换域中的系数进行噪声抑制操作，目的是提高局部特征信息提取的可靠性。变换域中噪声抑制方法是比较成熟的技术，这里不再赘述。
此后进入步骤303，从噪声干扰抑制的结果中提取特征信息作为跨通道不变信息。特征信息提取主要是对变换域中的能量系数和图像纹理之间关系的挖掘。
在本发明的一个实例中，对亮度通道图像进行二维DCT变换，从变换后得到的系数矩阵的行可以提取出图像水平方向边缘信息，比如二维DCT系数矩阵第一行系数绝对值之和为Fv；从变换后得到的系数矩阵的列中可以提取出图像垂直方向边缘信息，比如系数矩阵第一列系数绝对值之和为Fh；从变换后得到的系数矩阵的稀疏程度可以提取出图像纹理信息，比如系数矩阵中非零系数的个数为F0。为了得到较好的特性信息提取，可采用更复杂的提取和分析方法。
步骤103的局部色度图像去噪可以采用与第二实施方式的步骤103相同的方法。
本发明的各方法实施方式均可以以软件、硬件、固件等方式实现。不管本发明是以软件、硬件、还是固件方式实现，指令代码都可以存储在任何类型的计算机可访问的存储器中（例如永久的或者可修改的，易失性的或者非易失性的，固态的或者非固态的，固定的或者可更换的介质等等）。同样，存储器可以例如是可编程阵列逻辑（Programmable
Array Logic，简称“PAL”） 、随机存取存储器（Random Access Memory，简称“RAM”）
、可编程只读存储器（Programmable Read Only Memory，简称“PROM”） 、只读存储器（Read-Only
Memory，简称“ROM”） 、电可擦除可编程只读存储器（Electrically Erasable Programmable
ROM，简称“EEPROM”） 、磁盘、光盘、数字通用光盘（Digital Versatile Disc，简称“DVD”）等等。
本发明第四实施方式涉及一种彩色图像去噪声的系统。图4是该彩色图像去噪声的系统的结构示意图。该彩色图像去噪声的系统包括：
图像分解单元，用于将彩色图像分解为一个亮度通道图像和N个色度通道图像，N≥2。
跨通道不变信息提取单元，用于从图像分解单元分解出的亮度通道图像中提取跨通道不变信息。跨通道不变信息包括但不限于各个通道间共有的图像信息（比如边缘纹理信息），以及可以通过简单运算得到的并且稳定的各通道的信息（比如噪声强度）等信息中的一种或某几种。
去噪声单元，用于根据跨通道不变信息提取单元所得的跨通道不变信息，对图像分解单元分解出的至少一个色度通道图像进行去噪声处理。在本发明的一个优选实例中，去噪声单元根据跨通道不变信息，对N个色度通道图像均进行去噪声处理。
本发明的各实施方式中，将色彩图像分为亮度图像和色度图像两大类通道图像，在色度图像去噪过程中利用了从亮度图像提取到的跨通道不变信息，并指导色度图像动态地采用不同的去噪策略抑制噪声。
第一实施方式是与本实施方式相对应的方法实施方式，本实施方式可与第一实施方式互相配合实施。第一实施方式中提到的相关技术细节在本实施方式中依然有效，为了减少重复，这里不再赘述。相应地，本实施方式中提到的相关技术细节也可应用在第一实施方式中。
本发明第五实施方式涉及一种彩色图像去噪声的系统。
第五实施方式在第四实施方式的基础上进行了改进，主要改进之处在于在跨通道不变信息提取单元中采用了包括预处理在内的高效跨通道不变信息提取方法。
具体地说，如图5所示，跨通道不变信息提取单元包括以下模块：
预处理模块，用于对亮度通道图像的局部图像进行局部数据预处理，该局部数据预处理用于抑制部分噪声的影响。亮度图像局部数据经过局部数据预处理，抑制部分噪声的影响，提高局部特征信息收集的准确度，该处理的繁简与否取决于特性信息收集和提取的精度需求，如精度要求不高，简单的高斯平滑滤波器即可，如果精度要求比较高，采用一个效果较好的滤波器实现预处理也未尝不可，比如双边滤波器等。
收集模块，用于对预处理模块的处理结果进行局部图像特征信息收集。收集模块可以使用soble算子、laplace算子、canny算子、Gabor滤波器等收集局部图像特征信息。
提取模块，用于根据收集模块输出的局部图像特征信息，计算表征局部特征的信息作为跨通道不变信息。
去噪声单元可以包括图6所示的模块：
局部块均值计算模块，用于计算色度通道图像中局部块均值V0。
局部块中心行均值计算模块，用于计算色度通道图像中局部块中心行均值V1。
局部块中心列均值计算模块，用于计算色度通道图像中局部块中心列均值V2。
跨通道不变信息融合去噪声模块，用于融合色度图像的局部数据信息和跨通道间的不变信息，综合完成色度图像去噪。在本发明的一个实例中，一种简单的融合方式如下：R
= Fv*V1 + Fh*V2 + Q(Fh， Fv)*V0，R为融合结果即去噪的结果，Q是Fh和Fv的函数，比如Q(Fh，Fv) = k -
max(Fh，Fv)；k是用于控制跨通道不变信息融合程度的因子。
局部块均值计算模块、局部块中心行均值计算模块和局部块中心列均值计算模块也可以不设置在去噪声单元中，而是设置在其它单元中，这种变化只是常规手段。
第二实施方式是与本实施方式相对应的方法实施方式，本实施方式可与第二实施方式互相配合实施。第二实施方式中提到的相关技术细节在本实施方式中依然有效，为了减少重复，这里不再赘述。相应地，本实施方式中提到的相关技术细节也可应用在第二实施方式中。
本发明第六实施方式涉及一种彩色图像去噪声的系统。
第六实施方式在第四实施方式的基础上进行了改进，主要改进之处在于：对亮度通道图像进行局部数据变换后，在变换域中提取跨通道不变信息。这种技术方案可以利用信号在变换域中表现出的不同特性，更有效准确地提取局部图像数据的边缘纹理等特征信息，从而更好地利用这些信息对色度通道图像进行去噪声处理。具体地说，如图7所示，跨通道不变信息提取单元包括以下模块：
变换模块，用于对亮度通道图像进行局部数据变换。
噪声干扰抑制模块，用于对变换模块的处理结果进行噪声干扰抑制。噪声干扰抑制模块主要是为了对局部变换域中的系数进行噪声抑制操作，目的是提高局部特征信息提取的可靠性。变换域中的噪声抑制方法是比较成熟的技术，这里不再赘述。
提取模块，用于从噪声干扰抑制模块的处理结果中提取特征信息作为跨通道不变信息。
第三实施方式是与本实施方式相对应的方法实施方式，本实施方式可与第三实施方式互相配合实施。第三实施方式中提到的相关技术细节在本实施方式中依然有效，为了减少重复，这里不再赘述。相应地，本实施方式中提到的相关技术细节也可应用在第三实施方式中。
本发明各实施方式中提到了一些跨通道不变信息的具体形式和一些具体的去噪方式，这只是为了便于理解和实施，其实并不必限于这些具体形式和去噪方式，只要是利用跨通道不变信息并结合色度图像局部数据信息的融合去噪都可以达到一定的效果，属于本发明要保护的范围。
需要说明的是，本发明各系统实施方式中提到的各单元或模块都是逻辑的单元或模块，在物理上，一个逻辑的单元或模块可以是一个物理的单元或模块，也可以是一个物理的单元或模块的一部分，还可以以多个物理的单元或模块的组合实现，这些逻辑的单元或模块本身的物理实现方式并不是最重要的，这些逻辑的单元或模块所实现的功能的组合是才解决本发明所提出的技术问题的关键。此外，为了突出本发明的创新部分，本发明上述各系统实施方式并没有将与解决本发明所提出的技术问题关系不太密切的单元或模块引入，这并不表明上述系统实施方式并不存在其它的单元或模块。
虽然通过参照本发明的某些优选实施方式，已经对本发明进行了图示和描述，但本领域的普通技术人员应该明白，可以在形式上和细节上对其作各种改变，而不偏离本发明的精神和范围。
Claims (15)
1 ．一种彩色图像去噪声的方法，其特征在于，包括以下步骤：
将彩色图像分解为一个亮度通道图像和 N 个色度通道图像， N ≥ 2 ；
从所述亮度通道图像提取跨通道不变信息；
根据所述跨通道不变信息，对至少一个所述色度通道图像进行去噪声处理。
2 ．根据权利要求 1
所述的彩色图像去噪声的方法，其特征在于，所述根据跨通道不变信息，对至少一个所述色度通道图像进行去噪声处理的步骤中，
根据所述跨通道不变信息，对所述 N 个色度通道图像均进行去噪声处理。
3 ．根据权利要求 1
所述的彩色图像去噪声的方法，其特征在于，所述跨通道不变信息可以是以下之一或其任意组合：
边缘信息、纹理信息、噪声强度。
4 ．根据权利要求 1 至 3
中任一项所述的彩色图像去噪声的方法，其特征在于，所述从亮度通道图像提取跨通道不变信息的步骤包括以下子步骤：
对所述亮度通道图像的局部图像进行局部数据预处理，该局部数据预处理用于抑制噪声的影响；
进行局部图像特征信息收集；
根据所述局部图像特征信息收集的结果，计算表征局部特征的信息作为所述跨通道不变信息。
5 ．根据权利要求 4
所述的彩色图像去噪声的方法，其特征在于，所述局部数据预处理可以是高斯平滑或以双边滤波器滤波。
6 ．根据权利要求 4
所述的彩色图像去噪声的方法，其特征在于，所述进行局部图像特征信息收集的步骤中，所收集的局部图像特征信息包括局部纹理强度信息和纹理方向信息。
7 ．根据权利要求 4
所述的彩色图像去噪声的方法，其特征在于，所述进行局部图像特征信息收集的步骤中，可以使用以下算子之一收集局部图像特征信息：
soble 算子、 laplace 算子、 canny 算子、 Gabor 滤波器。
8 ．根据权利要求 7
所述的彩色图像去噪声的方法，其特征在于，所述根据跨通道不变信息，对至少一个所述色度通道图像进行去噪声处理的步骤包括以下步骤：
计算色度通道图像的局部块均值 V0 、中心行均值 V1 和中心列均值 V2 ；
计算去噪结果 R = Fv*V1 + Fh*V2 + Q(Fh ， Fv)*V0 ，其中， Fv 为水平方向信息， Fh
垂直方向信息， Fv 和 Fh 是从所述亮度通道图像提取的跨通道不变信息， Q 是 Fh 和 Fv 的函数。
9 ．根据权利要求 8
所述的彩色图像去噪声的方法，其特征在于，所述从亮度通道图像提取跨通道不变信息的步骤中：
计算 soble 算子在局部数据区域水平方向的响应 Sh 的绝对值，作为提取出来的水平方向信息 Fh
；
计算 soble 算子在局部数据区域垂直方向的响应 Sv 的绝对值，作为提取出来的垂直方向信息 Fv
。
10 ．根据权利要求 1 至 3
中任一项所述的彩色图像去噪声的方法，其特征在于，所述从亮度通道图像提取跨通道不变信息的步骤包括以下子步骤：
对所述亮度通道图像进行局部数据变换；
在所述局部数据变换所得的变换域中进行噪声干扰抑制；
从所述噪声干扰抑制的结果中提取特征信息作为所述跨通道不变信息。
11 ．一种彩色图像去噪声的系统，其特征在于，包括：
图像分解单元，用于将彩色图像分解为一个亮度通道图像和 N 个色度通道图像， N ≥ 2 ；
跨通道不变信息提取单元，用于从所述图像分解单元分解出的亮度通道图像中提取跨通道不变信息；
去噪声单元，用于根据所述跨通道不变信息提取单元所得的跨通道不变信息，对所述图像分解单元分解出的至少一个色度通道图像进行去噪声处理。
12 ．根据权利要求 11 所述的彩色图像去噪声的系统，其特征在于，所述去噪声单元根据所述跨通道不变信息，对所述 N
个色度通道图像均进行去噪声处理；
所述跨通道不变信息可以是以下之一或其任意组合：
边缘信息、纹理信息、噪声强度。
13 ．根据权利要求 11 或 12
所述的彩色图像去噪声的系统，其特征在于，所述跨通道不变信息提取单元包括以下模块：
预处理模块，用于对所述亮度通道图像的局部图像进行局部数据预处理，该局部数据预处理用于抑制噪声的影响；
收集模块，用于对所述预处理模块的处理结果进行局部图像特征信息收集；
提取模块，用于根据所述收集模块输出的局部图像特征信息，计算表征局部特征的信息作为所述跨通道不变信息。
14 ．根据权利要求 13
所述的彩色图像去噪声的系统，其特征在于，所述预处理模块可以是高斯平滑滤波器或双边滤波器；
所述收集模块可以使用以下算子之一收集局部图像特征信息：
soble 算子、 laplace 算子、 canny 算子、 Gabor 滤波器。
15 ．根据权利要求 11 或 12
所述的彩色图像去噪声的系统，其特征在于，所述跨通道不变信息提取单元包括以下模块：
变换模块，用于对所述亮度通道图像进行局部数据变换；
噪声干扰抑制模块，用于对所述变换模块的处理结果进行噪声干扰抑制；
提取模块，用于从所述噪声干扰抑制模块的处理结果中提取特征信息作为所述跨通道不变信息。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110081617.7A CN102156964B (zh) | 2011-03-31 | 2011-03-31 | 彩色图像去噪声的方法及其系统 |
| CN201110081617.7 | 2011-03-31 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2012129897A1 true WO2012129897A1 (zh) | 2012-10-04 |
Family
ID=44438446
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2011/080040 WO2012129897A1 (zh) | 2011-03-31 | 2011-09-22 | 彩色图像去噪声的方法及其系统 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN102156964B (zh) |
| WO (1) | WO2012129897A1 (zh) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103955903A (zh) * | 2014-05-09 | 2014-07-30 | 东南大学 | 一种权值窗口自适应非局部均值图像去噪方法 |
Families Citing this family (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102156964B (zh) * | 2011-03-31 | 2013-10-30 | 杭州海康威视数字技术股份有限公司 | 彩色图像去噪声的方法及其系统 |
| CN102750671B (zh) * | 2011-09-28 | 2017-08-25 | 新奥特（北京）视频技术有限公司 | 一种图像彩色噪声去除方法 |
| CN102497560B (zh) * | 2011-12-02 | 2014-04-02 | 浙江工商大学 | 一种去除复合视频解码图像色串的方法 |
| CN103049756A (zh) * | 2012-11-16 | 2013-04-17 | 健雄职业技术学院 | 基于cema算法和纹理块匹配修复技术的自动提取与去除彩色图像内文字的方法 |
| CN105447816B (zh) * | 2014-09-25 | 2019-01-08 | 杭州海康威视数字技术股份有限公司 | 一种多通道图像放大方法和装置 |
| CN105631836B (zh) * | 2014-11-06 | 2018-10-23 | 杭州海康威视数字技术股份有限公司 | 一种图像降噪方法和装置 |
| CN104749334A (zh) * | 2015-02-16 | 2015-07-01 | 燕山大学 | 基于模式识别的生物式异常水质评价系统设计方法 |
| CN107480676B (zh) * | 2017-07-28 | 2019-12-24 | 浙江大华技术股份有限公司 | 一种车辆颜色识别方法、装置和电子设备 |
| WO2019019772A1 (en) * | 2017-07-28 | 2019-01-31 | Zhejiang Dahua Technology Co., Ltd. | SYSTEMS AND METHODS FOR IMAGE PROCESSING |
| CN107451976B (zh) * | 2017-07-28 | 2019-09-17 | 浙江大华技术股份有限公司 | 一种图像处理方法及装置 |
| CN108810320B (zh) * | 2018-06-01 | 2020-11-24 | 深圳市商汤科技有限公司 | 图像质量提升方法和装置 |
| CN110599406B (zh) | 2019-03-18 | 2022-05-03 | 上海立可芯半导体科技有限公司 | 一种图像增强方法及装置 |
| CN113066033B (zh) * | 2021-04-19 | 2023-11-17 | 智领高新科技发展(北京)有限公司 | 一种彩色图像的多阶段去噪系统及方法 |
| CN117376485B (zh) * | 2023-10-23 | 2024-04-23 | 上海为旌科技有限公司 | 一种yuv联合去除图像色度噪声的方法和滤波器 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101103374A (zh) * | 2005-01-11 | 2008-01-09 | 伊斯曼柯达公司 | 稀疏分布的彩色数字图像的噪声清除 |
| CN101334836A (zh) * | 2008-07-30 | 2008-12-31 | 电子科技大学 | 一种融合色彩、尺寸和纹理特征的车牌定位方法 |
| CN101562692A (zh) * | 2008-04-02 | 2009-10-21 | 英属开曼群岛商恒景科技股份有限公司 | 噪声降低装置及方法 |
| CN102156964A (zh) * | 2011-03-31 | 2011-08-17 | 杭州海康威视软件有限公司 | 彩色图像去噪声的方法及其系统 |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1213592C (zh) * | 2001-07-31 | 2005-08-03 | 佳能株式会社 | 采用自适应二值化的图象处理方法和设备 |
| CN100559836C (zh) * | 2007-09-20 | 2009-11-11 | 宝利微电子系统控股公司 | 一种数字图像降噪方法 |
| CN101489034B (zh) * | 2008-12-19 | 2010-12-01 | 四川虹微技术有限公司 | 一种视频图像噪声估计与去除方法 |
| CN101957984B (zh) * | 2010-08-30 | 2012-03-28 | 西安电子科技大学 | 基于非局部萎缩因子参数估计的图像去噪方法 |
| CN101950414B (zh) * | 2010-09-02 | 2012-07-04 | 西安电子科技大学 | 自然图像非局部均值去噪方法 |
-
2011
- 2011-03-31 CN CN201110081617.7A patent/CN102156964B/zh active Active
- 2011-09-22 WO PCT/CN2011/080040 patent/WO2012129897A1/zh active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101103374A (zh) * | 2005-01-11 | 2008-01-09 | 伊斯曼柯达公司 | 稀疏分布的彩色数字图像的噪声清除 |
| CN101562692A (zh) * | 2008-04-02 | 2009-10-21 | 英属开曼群岛商恒景科技股份有限公司 | 噪声降低装置及方法 |
| CN101334836A (zh) * | 2008-07-30 | 2008-12-31 | 电子科技大学 | 一种融合色彩、尺寸和纹理特征的车牌定位方法 |
| CN102156964A (zh) * | 2011-03-31 | 2011-08-17 | 杭州海康威视软件有限公司 | 彩色图像去噪声的方法及其系统 |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103955903A (zh) * | 2014-05-09 | 2014-07-30 | 东南大学 | 一种权值窗口自适应非局部均值图像去噪方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102156964B (zh) | 2013-10-30 |
| CN102156964A (zh) | 2011-08-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2012129897A1 (zh) | 彩色图像去噪声的方法及其系统 | |
| JP2004522372A (ja) | 時空間適応的雑音除去／高画質復元方法及びこれを応用した高画質映像入力装置 | |
| Nithyananda et al. | Review on histogram equalization based image enhancement techniques | |
| JP4290193B2 (ja) | 画像処理装置 | |
| CN111161167B (zh) | 基于中通道补偿和自适应大气光估计的单幅图像去雾方法 | |
| WO2000016265A1 (en) | A general image enhancement framework | |
| US7526099B2 (en) | Motion picture data processing device, method, and program | |
| JP2012018669A (ja) | デジタル入力画像を処理する方法 | |
| CN101562692A (zh) | 噪声降低装置及方法 | |
| CN111612705B (zh) | 一种增强ocr性能的背景图像消除方法 | |
| KR100869134B1 (ko) | 이미지 처리장치 및 방법 | |
| CN115578294B (zh) | 图像增强方法、装置、设备及存储介质 | |
| KR102183518B1 (ko) | 이미지 신호로부터 계단 아티팩트들을 감소시키는 이미지 프로세싱 장치 | |
| KR100839922B1 (ko) | 적응적 평균값 필터를 이용한 컬러 잡음 제거 장치 및 방법 | |
| JP2004236185A (ja) | 信号処理装置および方法、記録媒体、並びにプログラム | |
| JP5227906B2 (ja) | 映像記録システム | |
| CN109167892B (zh) | 一种视频图像细节增强方法与系统 | |
| CN113379629A (zh) | 卫星图像去噪方法、装置、计算机设备和存储介质 | |
| Xiang et al. | Hazy image enhancement based on the full-saturation assumption | |
| KR20140054733A (ko) | 영상 처리 방법 및 장치 | |
| Mittal et al. | An efficient video enhancement method using LA* B* analysis | |
| KR101022832B1 (ko) | Ｃｃｄ센서의 스미어 열화 개선장치 및 방법 | |
| Date et al. | Low light video enhancement: A Survey | |
| Lu et al. | A multiresolution approach to recovering colors and details of clipped image regions | |
| CN116977221A (zh) | 一种针对矿井重粉尘环境的图像复原方法 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11862288 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 11862288 Country of ref document: EP Kind code of ref document: A1 |
|
| 32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 21.05.2014) |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 11862288 Country of ref document: EP Kind code of ref document: A1 |