WO2017016295A1 - 一种自适应视频空域去噪方法及装置 - Google Patents
一种自适应视频空域去噪方法及装置 Download PDFInfo
- Publication number
- WO2017016295A1 WO2017016295A1 PCT/CN2016/083056 CN2016083056W WO2017016295A1 WO 2017016295 A1 WO2017016295 A1 WO 2017016295A1 CN 2016083056 W CN2016083056 W CN 2016083056W WO 2017016295 A1 WO2017016295 A1 WO 2017016295A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- pixel
- current
- denoising
- adjacent
- pixel point
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/76—Television signal recording
- H04N5/91—Television signal processing therefor
- H04N5/911—Television signal processing therefor for the suppression of noise
Definitions
- the present application relates to the field of video technologies, and in particular, to a method and apparatus for an adaptive video spatial domain denoising method.
- Video denoising methods can be basically divided into time domain denoising, spatial domain denoising, and time domain plus spatial domain denoising. Most of the current denoising methods need to set the denoising intensity in advance, and then denoise each pixel of the video according to the same denoising intensity set. This processing can achieve denoising effects for noisy video, but for video with varying noise levels or no noise, the details within the processed video frame will be greatly lost. Therefore, it is necessary to find a denoising method that can automatically adjust the denoising intensity according to the noise intensity of the video frame.
- the invention proposes an adaptive spatial domain video denoising method, which can automatically set the denoising intensity according to the noise intensity of each pixel in the video frame to complete the denoising. This method avoids the loss of detail caused by video without noise pixel while ensuring effective denoising of noise pixel points.
- the present application provides an adaptive video spatial domain denoising method and apparatus for dynamically denoising processing according to the noise intensity of each pixel point in a video frame.
- an adaptive video spatial domain denoising method includes:
- a computer readable recording medium having recorded thereon a program for executing the above method.
- an adaptive video spatial denoising apparatus includes:
- a pixel value obtaining module configured to respectively acquire pixel values of each pixel point at the same position in the current frame and the adjacent frame, and also respectively acquire adjacent to the four sides of the current pixel point in the current frame The pixel value of the pixel;
- a normalization processing module configured to perform normalization processing on the acquired pixel values of the current frame and each pixel at the same position in the adjacent previous frame;
- a noise intensity calculation module configured to normalize the current frame acquired by the pixel value acquisition module and the pixel value of each pixel at the same position in the adjacent previous frame, and also used for Calculating a noise intensity of the current pixel point according to a pixel value of a current pixel point in the current frame and a pixel value of a pixel point in the adjacent previous frame that is the same as a position of the current pixel point;
- An adaptive spatial domain denoising module configured to perform adaptive spatial denoising on the current pixel point according to the noise intensity, a pixel value of the current pixel point, and a pixel value of an adjacent pixel on the upper, lower, left, and right sides .
- Figure 1 is a flow chart of the first application
- FIG. 2 is a schematic diagram of pixel points at the same position of the adjacent previous frame and the current frame of the present application;
- FIG. 3 is a flow chart of the second application
- FIG. 4 is a schematic diagram of a noise intensity function corresponding to a pixel value difference of pixels at the same position between two adjacent frames in the present application;
- Figure 5 is a flow chart of the third application
- FIG. 6 is a schematic diagram of a pixel point of a current pixel point of the present application and four sides of the upper, lower, left, and right sides thereof;
- Figure 7 is a block diagram of the apparatus of the fourth application.
- the adaptive video denoising method of the present invention mainly includes the following steps:
- Step 101 respectively acquire pixel values of all pixels at the same position in the current frame and the adjacent previous frame;
- the pixel point in the current frame is P(i, j)
- the pixel point at the same position in the adjacent previous frame is P'(i, j)
- i, j is a pixel point.
- the acquisition in this step is performed on all the pixels in the video frame.
- Step 102 normalize the pixel values of each pixel point at the same position in the current frame and the adjacent frame.
- Step 103 Calculate according to the pixel value of the current pixel in the current frame after the normalization process and the pixel value of the pixel in the adjacent previous frame that is the same as the position of the current pixel. Describe the noise intensity of the current pixel;
- Step 104 respectively acquire pixel values of adjacent pixel points on the upper, lower, left, and right sides of the current pixel point in the current frame.
- Step 105 Perform adaptive spatial denoising on the current pixel point according to the noise intensity, the pixel value of the current pixel point, and the pixel value of the adjacent pixel points on the upper, lower, left, and right sides.
- the pixel value of the current pixel in the current frame after the normalization process and the position of the current pixel in the adjacent previous frame after the normalization process are the same
- the pixel value of the pixel is used to calculate the noise intensity of the current pixel, and further includes the following steps:
- Step 201 normalize the obtained pixel values
- Normalization is a way of simplifying computations, where a dimensional expression is transformed into a dimensionless expression that becomes a scalar. In this step, the obtained pixel value P(i, j) is normalized so that 0 ⁇ P ⁇ 1.
- V(i,j) is the result of the normalization calculation
- P(i,j) is the pixel value of each current pixel
- 255 is the maximum value of the pixel value
- 0 is the minimum value of the pixel value.
- Step 202 Calculate the absolute value of the difference between the pixel value of the pixel point after the normalization process and the pixel position of the adjacent pixel in the adjacent previous frame after the normalization process value.
- the occurrence of noise in the video is random, that is, the position at which the noise occurs is random between two adjacent video frames.
- the pixel value of each pixel at the same position of two adjacent frames does not change much. Therefore, there is a certain correspondence between the absolute value of the pixel value difference between the two adjacent video frames and the noise intensity.
- V(i,j) is the pixel value of the current pixel after the normalization process
- V'(i,j) is normalized
- the pixel values of the pixels in the adjacent previous frame that are the same as the position of the current pixel point, m and n are constant and are empirical values, and are preset according to the degree of denoising.
- Equation 2 The calculation method of noise intensity is shown in Equation 2:
- Equation 2 L(i,j) is the noise intensity, V' and V represent two two-dimensional matrices, and V' is the normalized pixel value of all the pixels on the previous video frame, V is The normalized pixel value of all the pixels on the current video frame, where m and n are constants, all of which are empirical values, and are adjusted according to the degree of denoising. After testing, the value of n ranges from 0.80 to 0.99. The adaptive denoising effect is optimal.
- the absolute value of the difference of the pixel values is approximately Gaussian with the noise intensity, when the absolute value of the difference of the pixel values is smaller than the first threshold or the difference of the pixel values.
- the noise intensity calculated by Equation 1 is approximately zero, indicating that the current pixel point is not noise and there is no screen switching between the current video frame and the previous video frame, wherein the first threshold Less than the second threshold.
- the pixel values of adjacent pixel points on the upper, lower, left, and right sides of the current pixel point in the current frame are respectively acquired, according to the noise intensity, the pixel value of the current pixel point, and the up, down, left, and right directions.
- the pixel values of the adjacent pixels on the four sides perform adaptive spatial denoising on the current pixel, and further include the following steps:
- Step 301 Acquire pixel values of adjacent pixel points on the upper, lower, left, and right sides of the current pixel point in the current frame.
- the current pixel point pixel value is P(i, j), and the pixel of the adjacent pixel point on the left side
- the value is P(i-1,j)
- the pixel value of the adjacent pixel on the right side is P(i+1,j)
- the pixel value of the pixel adjacent to the upper side is P(i,j-1)
- the pixel value of the pixel adjacent to the lower side is P(i, j+1).
- Equation 3 The calculation method of the denoising weight of the current pixel point is as shown in Equation 3:
- Equation 3 w m is the denoising weight of the current pixel point, and x, y are empirical values and are set according to the noise intensity.
- the noise intensity L(i,j) is greater than a certain threshold, the denoising weight of the current pixel point is reduced by decreasing x, y, thereby reducing the denoising weight of the noise point to Achieve better denoising effect.
- Step 303 Calculate denoising weights of adjacent pixel points of the upper, lower, left, and right sides according to pixel values of the current pixel point and pixel values of adjacent pixel points of the upper, lower, left, and right sides.
- Equation 4 The denoising weights of the adjacent pixels on the upper, lower, left, and right sides are calculated by Equation 4, and Equation 4 is as follows:
- Equation 4 is a normal distribution of deformation
- f x is a normal distribution function
- x is a random variable
- ⁇ is a normal distribution standard deviation.
- the difference between the pixel value of the current pixel point and the pixel value of the adjacent pixel points on the upper, lower, left, and right sides is a random variable x, and is calculated according to a preset standard deviation ⁇ .
- the specific calculation method is as shown in the following formula:
- Step 304 Performing weighted averaging according to the pixel value of the current pixel point, the denoising weight of the current pixel point, the pixel value of the adjacent pixel points of the upper, lower, left, and right sides, and the denoising weight of the adjacent pixel point a value that replaces the pixel value of the current pixel point with the average value.
- the denoising weight of the current pixel point is multiplied by the pixel value of the current pixel point, and the denoising weight of the adjacent pixel points of the upper, lower, left, and right sides is multiplied by the adjacent pixel points of the upper, lower, left, and right sides.
- a pixel value as a weighted summation result the sum of the denoising weight of the current pixel point and the denoising weight of the adjacent pixel points of the upper, lower, left, and right sides as the base of the weighted averaging, with the weighting
- the result of the sum is divided by the base to obtain the result of the weighted averaging instead of the pixel value of the current pixel.
- N(i,j) [w m *P(i,j)+w l *P(i-1,j)+w r *P(i+1,j)+w t *P(i,j -1+wb*Pi,j+1]/(wm+wl+wr+wt+wb) Equation 6
- N(i, j) is an average value obtained by weighted averaging, and N(i, j) is used instead of the pixel value of the current pixel point. This step is performed by traversing all the noise points on each current video frame, and details are not described herein.
- the invention can realize the adaptive adjustment of the denoising intensity by calculating the noise intensity of the video frame, so that the video with the noise intensity variation or no noise can retain its intra-frame details, which is more conducive to improving the quality of the video and the viewing experience of the viewer.
- the present invention relates to an adaptive spatial video denoising apparatus, including a pixel value acquisition module 701, a normalization processing module 702, a noise intensity calculation module 703, an adaptive spatial domain denoising module 704, and a weight. Calculation module 705.
- the pixel value obtaining module 701 is configured to respectively acquire pixel values of each pixel point at the same position in the current frame and the adjacent frame, and also respectively acquire adjacent to the four sides of the current pixel point in the current frame.
- the normalization processing module 702 is configured to perform normalization processing on the acquired pixel values of the current frame and each pixel at the same position in the adjacent previous frame;
- the noise intensity calculation module 703 is configured to: according to the normalized pixel value of the current pixel in the current frame and the pixel value of the pixel in the adjacent previous frame that is the same as the current pixel position Calculating a noise intensity of the current pixel point;
- the adaptive spatial domain denoising module 704 is configured to perform adaptive spatial denoising on the current pixel point according to the noise intensity, the pixel value of the current pixel point, and the pixel value of the adjacent pixel points of the upper, lower, left, and right sides. .
- the noise intensity calculation module 703 is further configured to calculate, by the normalized processing, the current pixel point and the pixel position of the adjacent previous frame that is the same as the current pixel point in the normalized processing.
- the pixel values of the same pixel point as the current pixel point, m and n are constant and are empirical values, and are preset according to the degree of denoising.
- the adaptive spatial domain denoising module 704 is further configured to: according to the pixel value of the current pixel point, the denoising weight of the current pixel point, the pixel value of the adjacent pixel points of the upper, lower, left, and right sides, and the upper, lower, left, and right sides.
- the denoising weights of adjacent pixel points are weighted and averaged to obtain an average value, and the average value is substituted for the pixel value of the current pixel point.
- the weight calculation module 705 is further configured to calculate a denoising weight of the adjacent pixel points of the upper, lower, left, and right sides according to pixel values of the current pixel point and pixel values of adjacent pixel points of the upper, lower, left, and right sides, respectively. .
- the pixel values of each pixel at the same position in the current frame and the adjacent frame are respectively obtained.
- the acquired phase of the current pixel in the current frame is up, down, left, and right.
- N(i,j) (50*2.420+60*0.801+60*0.801+60*0.801+60*0.801)/(2.420+0.801+0.801+0.801) ⁇ 56.
- the pixel value 56 obtained by the denoising process is closer to the current frame than the pixel value 50 before the replacement.
- the adaptive video spatial domain denoising method and apparatus provided by the present application can dynamically adjust the denoising intensity according to the noise intensity of each pixel in the video frame to complete the denoising process.
- the present invention is capable of adaptively determining by noise intensity, thereby avoiding the details of noise-free video frames while ensuring effective denoising of noisy video frames. loss.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Picture Signal Circuits (AREA)
- Image Processing (AREA)
Abstract
本发明实施例提供一种自适应视频空域去噪的方法，通过获取当前帧与其相邻帧前一中相同位置处的每个像素点的像素值从而计算所述当前像素点的噪声强度；分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值，根据所述噪声强度、所述当前像素点的像素值以及所述上下左右四侧的相邻像素点的像素值计算所述当前像素点以及所述上下左右四侧的相邻像素点的去噪权重，利用加权求平均求得的值代替所述当前像素点的像素值从而实现对所述当前像素点的自适应空域去噪，在去噪的同时最大程度的保留了画面的细节。
Description
交叉引用
本申请引用于2015年7月24日递交的名称为“一种自适应视频空域去噪方法及装置”的第201510440941.1号中国专利申请，其通过引用被全部并入本申请。
本申请涉及视频技术领域，尤其涉及一种自适应视频空域去噪方法的方法及装置。
随着数字视频应用的迅猛发展，在数字视频系统中，视频的采集、传输、编码、解码等过程会不可避免地引入各种噪声，噪声的存在不但严重影响了视频主观视觉质量，而且会影响视频的后续处理，例如编码、转码等。因此，伴随着数字视频的广泛应用，迫切需要有高效的视频去噪方法。
视频的去噪方法基本上可以分为时间域去噪、空间域去噪和时间域加空间域去噪等类型。目前的去噪方法大多需要预先设置好去噪强度，之后对视频的每个像素点按照设置好的相同的去噪强度进行去噪处理。这样处理对于有噪声的视频能够达到去噪效果，但对于噪声强度发生变化或者没有噪声的视频，处理后的视频帧内的细节将大大损失。因此，寻找一种能够根据视频帧噪声强度自动调节去噪强度的去噪方法是十分必要的。
本发明提出了一种自适应空域视频去噪方法，能够根据视频帧内每个像素点的噪声强度自动设置去噪强度完成去噪。该方法在保证对噪声像素点有效去噪的同时，避免了对没有噪声像素点视频造成的细节损失。
发明内容
本申请提供一种自适应视频空域去噪方法及装置，用以根据视频帧内每个像素点的噪声强度动态调节去噪强度完成去噪处理。
为达到上述目的，本申请实施例采用如下技术方案：
第一方面，一种自适应视频空域去噪方法，包括：
分别获取当前帧与其相邻前一帧中相同位置处的所有像素点的像素值并对获取到的像素值进行归一化处理；
根据归一化处理后的所述当前帧中的当前像素点的像素值及所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值来计算所述当前像素点的噪声强度；
分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值；
根据所述噪声强度、所述当前像素点的像素值以及所述上下左右四侧的相邻像素点的像素值对所述当前像素点进行自适应空域去噪。
第二方面，一种在其上记录有用于执行上述方法的程序的计算机可读记录介质。
第三方面，一种自适应视频空域去噪装置，包括：
像素值获取模块：用于分别获取当前帧与其相邻帧中相同位置处的每个像素点的像素值，还用于分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值；
归一化处理模块：用于对获取到的所述当前帧与所述相邻前一帧中相同位置处的每个像素点的像素值进行归一化处理；
噪声强度计算模块：用于将所述像素值获取模块获取到的所述当前帧与所述相邻前一帧中相同位置处的每个像素点的像素值进行归一化处理，还用于根据所述当前帧中的当前像素点的像素值及所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值来计算所述当前像素点的噪声强度；
自适应空域去噪模块：用于根据所述噪声强度、所述当前像素点的像素值以及所述上下左右四侧的相邻像素点的像素值对所述当前像素点进行自适应空域去噪。
附图概述
为了更清楚地说明本申请或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本申请一的流程图；
图2是本申请的相邻前一帧与当前帧相同位置处像素点的示意图；
图3是本申请二的流程图；
图4是本申请的相邻两帧之间相同位置像素点的像素值差值对应的噪声强度函数示意图；
图5是本申请三的流程图；
图6是本申请的当前像素点与其上下左右四侧的像素点的示意图；
图7是本申请四的装置结构图。
本申请的较佳实施方式
为使本申请的目的、技术方案和优点更加清楚，下面将结合本申请中的附图，对本申请中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
实施例一
如图1所示，本发明的自适应视频去噪方法主要包括以下步骤：
步骤101：分别获取当前帧与其相邻前一帧中相同位置处的所有像素点的像素值；
如图2示，所述当前帧中的像素点为P(i，j)，所述相邻前一帧中相同位置处的像素点为P’(i，j)，i，j是像素点在所在帧内的坐标，本步骤中的获取对视频帧内的所有像素点均遍历执行。
步骤102：对获取到的所述当前帧与其相邻帧中相同位置处的每个像素点的像素值进行归一化处理；
步骤103：根据归一化处理后的所述当前帧中的当前像素点的像素值及所述相邻前一帧中的与所述当前像素点的位置相同的像素点的像素值来计算所述当前像素点的噪声强度；
步骤104：分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值；
步骤105：根据所述噪声强度、所述当前像素点的像素值以及所述上下左右四侧的相邻像素点的像素值对所述当前像素点进行自适应空域去噪。
实施例二
如图3所示，根据归一化处理后的所述当前帧中的当前像素点的像素值及归一化处理后的所述相邻前一帧中的与所述当前像素点的位置相同的像素点的像素值来计算所述当前像素点的噪声强度，进一步包括如下步骤：
步骤201：对获取到的像素值进行归一化处理；
归一化是一种简化计算的方式，即将有量纲的表达式，经过变换，化为无量纲的表达式，成为标量。本步骤中，对获取到的像素值P(i，j)进行归一化处理，使得0≤P≤1。
归一化计算的具体公式如下：
公式1中，V(i，j)是归一化计算的结果，P(i，j)是每个当前像素点的像素值，255是像素值的最大值，0是像素值的最小值。
步骤202：计算归一化处理后的所述当前像素点与归一化处理后的所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值的差值的绝对值。
视频中的噪声出现是随机的，即噪声出现的位置在前后相邻两视频帧之间是随机的。在没有噪声且画面无切换的情况下，相邻两帧相同位置处的每个像素点的像素值变化不大。因此前后相邻两视频帧相同位置的像素值差值的绝对值与噪声强度之间存在一定的对应关系。
步骤203：使用公式L(i，j)＝(m*(1-|V’(i，j)-V(i，j)|))n*|V’(i，j)-V(i，j)|计算所述当前像素点的噪声强度，其中V(i，j)是归一化处理后的所述当前像素点的像素值，V’(i，j)是归一化处理后的所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值，m和n是常数且均为经验值，根据去噪程度进行预先设置。
噪声强度的计算方法如公式2所示：
L(i，j)＝(m*(1-|V’(i，j)-V(i，j)|))n*|V’(i，j)-V(i，j)| 公式2
公式2中，L(i，j)是所述噪声强度，V’和V代表的是两个二维矩阵，V’是前一视频帧上的所有像素点的归一化像素值，V是当前视频帧上的所有像素点的归一化像素值，其中，m、n为常数，均为经验值，根据去噪程度进行调整，经测试研究，n的取值范围在0.80～0.99之间时，自适应的去噪效果最优。
如图4的示意，所述像素值的差值的绝对值与所述噪声强度近似成高斯分布，当所述像素值的差值的绝对值小于第一阈值或所述像素值的差值的绝对值大于第二阈值时，通过公式1计算出的所述噪声强度近似为零，说明当前像素点不是噪声且当前视频帧和前一视频帧之间无画面切换，其中，所述第一阈值小于第二阈值。
实施例三
如图5所示，分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值，根据所述噪声强度、所述当前像素点的像素值以及所述上下左右四侧的相邻像素点的像素值对所述当前像素点进行自适应空域去噪，进一步包括如下步骤：
步骤301：分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值。
如图6所示，所述当前像素点像素值为P(i，j)，左侧相邻的像素点的像素
值为P(i-1，j)，右侧相邻的像素点的像素值为P(i+1，j)，上侧相邻的像素点的像素值为P(i，j-1)，下侧相邻的像素点的像素值为P(i，j+1)。
步骤302：根据公式wm＝x+y*L(i，j)计算所述当前像素点的去噪权重，其中x和y是经验值，根据所述当前像素点的所述噪声强度进行调整。
所述当前像素点的去噪权重的计算方法如公式3所示：
wm＝x+y*L(i，j) 公式3
公式3中，wm是当前像素点的去噪权重，x，y是经验值且根据所述噪声强度进行设置。当所述噪声强度L(i，j)大于某一特定的阈值时，通过减小x，y来减小所述当前像素点的所述去噪权重，从而减小噪声点的去噪权重以达到较好的去噪效果。
步骤303：根据所述当前像素点的像素值分别与所述上下左右四侧的相邻像素点的像素值计算所述上下左右四侧的相邻像素点的去噪权重。
所述上下左右四侧的相邻像素点的去噪权重以公式4进行计算，公式4如下所示：
公式4是正态分布的变形，fx是正态分布函数，x是随机变量，σ为正态分布标准差。本发明中，以所述当前像素点的像素值与所述上下左右四侧的相邻像素点的像素值的差值为随机变量x，根据预设的标准差σ进行计算。具体计算方法如以下公式所示：
xl＝[P(i-1，j)-P(i，j)
xr＝P(i+1，j)-P(i，j)
xt＝P(i，j-1)-P(i，j)
xb＝P(i，j+1)-P(i，j)
公式4中，xl、xr、xt、xb分别是所述当前像素点的像素值与所述上下左右四侧的相邻像素点的像素值的差值，wl是所述左侧相邻的像素点的去噪权重，wr是所述右侧相邻的像素点的去噪权重，wt是所述上侧相邻的像素点的去噪权重，wb是所述下侧相邻的像素点的去噪权重，σ是预设的标准差，一般取σ＝15。
步骤304：根据所述当前像素点的像素值、当前像素点的去噪权重、所述上下左右四侧的相邻像素点的像素值、相邻像素点的去噪权重进行加权求平均得到平均值，以所述平均值代替所述当前像素点的像素值。
所述当前像素点的去噪权重乘以所述当前像素点的像素值，再加所述上下左右四侧的相邻像素点的去噪权重乘以所述上下左右四侧的相邻像素点的像素值作为加权求和结果，以所述当前像素点的去噪权重与所述上下左右四侧的相邻像素点的去噪权重的和作为所述加权求平均的基数，以所述加权求和结果除以所述基数得到所述加权求平均的结果来代替所述当前像素点的像素值。
加权求平均的具体计算如下公式所示：
N(i，j)＝[wm*P(i，j)+wl*P(i-1，j)+wr*P(i+1，j)+wt*P(i，j-1+wb*Pi，j+1]/(wm+wl+wr+wt+wb) 公式6
公式6中，N(i，j)是加权求平均得到的平均值，用N(i，j)来代替所述当前像素点的像素值。本步骤对每一当前视频帧上的所有噪声点均遍历执行，此处不赘述。
本发明通过对视频帧噪声强度的计算能够实现去噪强度的自适应调节，从而对于噪声强度变化或者没有噪声的视频能够保留其帧内细节，更利于提升视频的质量以及观众的观看体验。
实施例四
如图7所示，本发明涉及到的一种自适应空域视频去噪装置，包括像素值获取模块701、归一化处理模块702、噪声强度计算模块703、自适应空域去噪模块704、权重计算模块705。
像素值获取模块701用于分别获取当前帧与其相邻帧中相同位置处的每个像素点的像素值，还用于分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值；
归一化处理模块702：用于对获取到的所述当前帧与所述相邻前一帧中相同位置处的每个像素点的像素值进行归一化处理；
噪声强度计算模块703用于根据归一化处理后的所述当前帧中的当前像素点的像素值及所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值来计算所述当前像素点的噪声强度；
自适应空域去噪模块704用于根据所述噪声强度、所述当前像素点的像素值以及所述上下左右四侧的相邻像素点的像素值对所述当前像素点进行自适应空域去噪。
所述噪声强度计算模块703进一步用于计算归一化处理后的所述当前像素点与归一化处理后的所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值的差值的绝对值，根据如下公式L(i，j)＝(m*(1-|V’(i，j)-Vi，j)n*V’i，j-Vi，j计算所述当前像素点的噪声强度，其中Vi，j是归一化处理后的所述当前像素点的像素值，V’(i，j)是归一化处理后的所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值，m和n是常数且均为经验值，根据去噪程度进行预先设置。
所述自适应空域去噪模块704进一步用于根据所述当前像素点的像素值、当前像素点的去噪权重、所述上下左右四侧的相邻像素点的像素值、上下左右四侧的相邻像素点的去噪权重进行加权求平均得到平均值，以所述平均值代替所述当前像素点的像素值。
所述自适应空域去噪模704块进一步包括权重计算模块705，所述权重计算模块704用于计算所诉当前像素点的去噪权重与所述上下左右四侧相邻的像素点的去噪权重，根据公式wm＝x+y*L(i，j)计算所述当前像素点的去噪权重，其中x和y是经验值，根据所述当前像素点的所述噪声强度进行调整；
所述权重计算模块705进一步用于根据所述当前像素点的像素值分别与所述上下左右四侧的相邻像素点的像素值计算所述上下左右四侧的相邻像素点的去噪权重。
应用实例
本实施例将结合实际应用场景来进一步阐述本发明。
首先分别获取当前帧与其相邻帧中相同位置处的每个像素点的像素值，本实施例中假设所述当前帧中的所述当前像素值为P(i，j)＝50，归一化值为
所述相邻前一帧中相同位置处的像素点的像素值为P’(i，j)＝60，归一化值为
根据获取到的像素值利用公式1计算所述当前像素点的噪声强度，本实施例中，预设m＝2，n＝0.9，则
L(i，j)＝(2*(1-|0.235-0.196|))0.9*|0.235-0.196|≈0.070。
分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值，本实施例中，假设获取到的所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值分别如下：P(i，j-1)＝60，P(i，j+1)＝60，P(i-1，j＝60Pi+1，j＝60。
根据公式3计算所述当前像素点的去噪权重，本实施例中常数x＝2，常数y＝6，wm＝2+6*L(i，j)＝2+6*0.07＝2.420。
根据公式5计算以所述上下左右四侧的相邻像素点的去噪权重，本实施例中σ＝15，取e＝2.71828：
根据公式6进行加权求平均得到平均值，以所述平均值代替所述当前像素点的像素值：
N(i，j)＝(50*2.420+60*0.801+60*0.801+60*0.801+60*0.801)/(2.420+0.801+0.801+0.801+0.801)≈56。
用计算出的56作为新的像素值代替获取到的所述当前像素点的像素值，与替换前的像素值50相比，通过去噪处理得到的像素值56更接近所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值。
以上所描述的装置实施例仅仅是示意性的，可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性的劳动的情况下，即可以理解并实施。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到各实施方式可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件。基于这样的理解，上述技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品可以存储在计算机可读存储介质中，如ROM/RAM、磁碟、光盘等，包括若干指令用以使得一台
计算机设备(可以是个人计算机，服务器，或者网络设备等)执行各个实施例或者实施例的某些部分所述的方法。
最后应说明的是：以上实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。
本申请提供的自适应视频空域去噪方法及装置，能够根据视频帧内每个像素点的噪声强度动态调节去噪强度完成去噪处理。对于噪声强度变化或者没有噪声的视频帧，本发明能够自适应地通过噪声强度进行判断，从而在保证对有噪声的视频帧进行有效去噪的同时，避免了对没有噪声的视频帧造成的细节损失。
Claims (11)
- 一种自适应视频空域去噪方法，其特征在于，包括：分别获取当前帧与其相邻前一帧中相同位置处的所有像素点的像素值并对获取到的像素值进行归一化处理；根据归一化处理后的所述当前帧中的当前像素点的像素值及所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值来计算所述当前像素点的噪声强度；分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值；根据所述噪声强度、所述当前像素点的像素值以及所述上下左右四侧的相邻像素点的像素值对所述当前像素点进行自适应空域去噪。
- 根据权利要求1所述的自适应空域视频去噪方法，其特征在于，计算所述当前像素点的噪声强度，进一步包括：使用如下公式L(i,j)＝(m*(1-|V’(i,j)-V(i,j)|))n*|V’(i,j)-V(i,j)|计算所述当前像素点的噪声强度，其中V(i,j)是归一化处理后的所述当前像素点的像素值，V’(i,j)是归一化处理后的所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值，m和n是常数且均为经验值，根据去噪程度进行预先设置。
- 根据权利要求1所述的自适应空域视频去噪方法，其特征在于，根据所述当前像素点的像素值、当前像素点的去噪权重、所述上下左右四侧的相邻像素点的像素值、上下左右四侧的相邻像素点的去噪权重进行加权求平均得到平均值，以所述平均值代替所述当前像素点的像素值。
- 根据权利要求3所述的自适应空域视频去噪方法，其特征在于，根据公式wm＝x+y*L(i,j)计算所述当前像素点的去噪权重，其中x和y是经验值，当所述噪声强度L(i,j)大于某一特定的阈值时，通过减小x和y减小所述当前像素点的去噪权重。
- 根据权利要求3所述的自适应空域视频去噪方法，其特征在于，所述上下左右四侧的相邻像素点的去噪权重以公式进行计算，其中，以所述上下左右四侧的相邻像素点的像素值与所述当前像素点的像素值的差值为随机变量x，σ是预设的标准差。
- 一种在其上记录有用于执行权利要求1-5中任一项所述方法的程序的计算机可读记录介质。
- 一种自适应视频空域去噪装置，其特征在于，包括以下模块：像素值获取模块，用于分别获取当前帧与其相邻帧中相同位置处的每个像素点的像素值，还用于分别获取所述当前帧中所述当前像素点上下左右四侧的相邻像素点的像素值；归一化处理模块，用于对获取到的所述当前帧与所述相邻前一帧中相同位置处的每个像素点的像素值进行归一化处理；噪声强度计算模块，用于将所述像素值获取模块获取到的所述当前帧与所述相邻前一帧中相同位置处的每个像素点的像素值进行归一化处理，还用于根据所述当前帧中的当前像素点的像素值及所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值来计算所述当前像素点的噪声强度；自适应空域去噪模块，用于根据所述噪声强度、所述当前像素点的像素值以及所述上下左右四侧的相邻像素点的像素值对所述当前像素点进行自适应空域去噪。
- 根据权利要求7所述的装置，其特征在于，所述噪声强度计算模块进一步用于：根据如下公式L(i,j)＝(m*(1-|V’(i,j)-V(i,j)|))n*|V’(i,j)-V(i,j)|计算所述当前像素点的噪声强度，其中V(i,j)是归一化处理后的所述当前像素点的像素值，V’(i,j)是归一化处理后的所述相邻前一帧中的与所述当前像素点位置相同的像素点的像素值，m和n是常数且均为经验值，根据去噪程度进行预先设置。
- 根据权利要求7所述的装置，其特征在于，所述自适应空域去噪模块进一步用于：根据所述当前像素点的像素值、当前像素点的去噪权重、所述上下左右四侧的相邻像素点的像素值、上下左右四侧的相邻像素点的去噪权重进行加权求平均得到平均值，以所述平均值代替所述当前像素点的像素值。
- 根据权利要求7所述的装置，其特征在于，所述自适应空域去噪模块进一步包括权重计算模块，所述权重计算模块用于计算所述当前像素点的去噪权重与所述上下左右四侧相邻的像素点的去噪权重，根据公式wm＝x+y*L(i,j)计算所述当前像素点的去噪权重，其中x和y是经验值，通过减小x和y减小所述当前像素点的去噪权重。
- 根据权利要求7所述的装置，其特征在于，所述权重计算模块进一步用于根据所述当前像素点的像素值分别与所述上下左右四侧的相邻像素点的像素值计算所述上下左右四侧的相邻像素点的去噪权重，所述上下左右四侧的相邻像素点的去噪权重以公式进行计算，以所述当前像素点的像素值与所述上下左右四侧的相邻像素点的像素值的差值为随机变量x，σ是预设的标准差。
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| RU2016136389A RU2016136389A (ru) | 2015-07-24 | 2016-05-23 | Способ и устройство для адаптивного шумоподавления в видео в пространственной области |
| US15/242,286 US20170024860A1 (en) | 2015-07-24 | 2016-08-19 | Method and device for adaptive spatial-domain video denoising |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510440941.1A CN105915761A (zh) | 2015-07-24 | 2015-07-24 | 一种自适应视频空域去噪方法及装置 |
| CN201510440941.1 | 2015-07-24 |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| US15/242,286 Continuation US20170024860A1 (en) | 2015-07-24 | 2016-08-19 | Method and device for adaptive spatial-domain video denoising |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2017016295A1 true WO2017016295A1 (zh) | 2017-02-02 |
Family
ID=56743961
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2016/083056 WO2017016295A1 (zh) | 2015-07-24 | 2016-05-23 | 一种自适应视频空域去噪方法及装置 |
Country Status (3)
| Country | Link |
|---|---|
| CN (1) | CN105915761A (zh) |
| RU (1) | RU2016136389A (zh) |
| WO (1) | WO2017016295A1 (zh) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070182862A1 (en) * | 2006-02-06 | 2007-08-09 | Li Xinghai Billy | Video display device, video encoder, noise level estimation module and methods for use therewith |
| US20070296871A1 (en) * | 2006-06-22 | 2007-12-27 | Samsung Electronics Co., Ltd. | Noise reduction method, medium, and system |
| US20080118179A1 (en) * | 2006-11-21 | 2008-05-22 | Samsung Electronics Co., Ltd. | Method of and apparatus for eliminating image noise |
| CN101658027A (zh) * | 2007-03-31 | 2010-02-24 | 索尼德国有限责任公司 | 用于图像帧的降噪方法和单元 |
| CN101742088A (zh) * | 2009-11-27 | 2010-06-16 | 西安电子科技大学 | 非局部均值空域时变视频滤波方法 |
| CN102281386A (zh) * | 2010-06-08 | 2011-12-14 | 中兴通讯股份有限公司 | 一种对视频图像进行自适应去噪的方法及装置 |
| CN104735300A (zh) * | 2015-03-31 | 2015-06-24 | 中国科学院自动化研究所 | 基于权重滤波的视频去噪装置及方法 |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102014240B (zh) * | 2010-12-01 | 2013-07-31 | 深圳市蓝韵实业有限公司 | 一种实时医学视频图像去噪方法 |
-
2015
- 2015-07-24 CN CN201510440941.1A patent/CN105915761A/zh active Pending
-
2016
- 2016-05-23 RU RU2016136389A patent/RU2016136389A/ru not_active Application Discontinuation
- 2016-05-23 WO PCT/CN2016/083056 patent/WO2017016295A1/zh active Application Filing
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070182862A1 (en) * | 2006-02-06 | 2007-08-09 | Li Xinghai Billy | Video display device, video encoder, noise level estimation module and methods for use therewith |
| US20070296871A1 (en) * | 2006-06-22 | 2007-12-27 | Samsung Electronics Co., Ltd. | Noise reduction method, medium, and system |
| US20080118179A1 (en) * | 2006-11-21 | 2008-05-22 | Samsung Electronics Co., Ltd. | Method of and apparatus for eliminating image noise |
| CN101658027A (zh) * | 2007-03-31 | 2010-02-24 | 索尼德国有限责任公司 | 用于图像帧的降噪方法和单元 |
| CN101742088A (zh) * | 2009-11-27 | 2010-06-16 | 西安电子科技大学 | 非局部均值空域时变视频滤波方法 |
| CN102281386A (zh) * | 2010-06-08 | 2011-12-14 | 中兴通讯股份有限公司 | 一种对视频图像进行自适应去噪的方法及装置 |
| CN104735300A (zh) * | 2015-03-31 | 2015-06-24 | 中国科学院自动化研究所 | 基于权重滤波的视频去噪装置及方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| RU2016136389A (ru) | 2018-03-15 |
| CN105915761A (zh) | 2016-08-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9196024B2 (en) | Method and apparatus for enhancing color | |
| JP5035029B2 (ja) | 信号処理装置および方法、並びにプログラム | |
| WO2020124873A1 (zh) | 图像处理方法 | |
| CN107437238B (zh) | 一种图像分块自适应递归降噪方法及装置 | |
| WO2009107197A1 (ja) | 画像処理装置、画像処理方法および画像処理プログラム | |
| KR20150090595A (ko) | 영상 처리 방법 및 장치 | |
| WO2017088391A1 (zh) | 视频去噪与细节增强方法及装置 | |
| US7555169B2 (en) | Method and a device for reducing image noises | |
| JP2009025862A (ja) | 画像処理装置、画像処理方法、画像処理プログラム及び画像表示装置 | |
| US8265419B2 (en) | Image processing apparatus and image processing method | |
| WO2016185708A1 (ja) | 画像処理装置、画像処理方法、および、記憶媒体 | |
| JP2017098604A (ja) | 映像品質推定装置、映像品質推定方法、及びプログラム | |
| JP2005150903A (ja) | 画像処理装置、ノイズ除去方法及びノイズ除去プログラム | |
| WO2017016295A1 (zh) | 一种自适应视频空域去噪方法及装置 | |
| JP2014112790A (ja) | 画像処理装置、画像処理方法及び画像処理プログラム | |
| WO2017036386A1 (zh) | 一种视频去噪方法及装置、终端、存储介质 | |
| US20170024860A1 (en) | Method and device for adaptive spatial-domain video denoising | |
| Sun et al. | Adaptive bilateral filter considering local characteristics | |
| Tanikawa et al. | Image restoration based on weighted average of multiple blurred and noisy images | |
| JP4913246B1 (ja) | エッジ強調方法またはエッジ強調度演算方法 | |
| TWI386868B (zh) | 使用內容調適懲罰函數的移動偵測方法 | |
| JP2012098861A (ja) | 画像処理装置、画像処理方法 | |
| JP2012129617A5 (ja) | 画像処理装置および方法並びにプログラム | |
| TWI680675B (zh) | 影像處理裝置與相關的影像處理方法 | |
| Xiao et al. | Reduction of mosquito noise based on the adaptive bilateral-filter |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| ENP | Entry into the national phase |
Ref document number: 2016136389 Country of ref document: RU Kind code of ref document: A |
|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 16829662 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: 16829662 Country of ref document: EP Kind code of ref document: A1 |