CN102236885A - Filter for reducing image noise and filtering method - Google Patents

Filter for reducing image noise and filtering method Download PDF

Info

Publication number
CN102236885A
CN102236885A CN2010101700703A CN201010170070A CN102236885A CN 102236885 A CN102236885 A CN 102236885A CN 2010101700703 A CN2010101700703 A CN 2010101700703A CN 201010170070 A CN201010170070 A CN 201010170070A CN 102236885 A CN102236885 A CN 102236885A
Authority
CN
China
Prior art keywords
window
difference
pixel
image noise
target
Prior art date
Application number
CN2010101700703A
Other languages
Chinese (zh)
Inventor
李东信
蒋俊成
Original Assignee
联咏科技股份有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 联咏科技股份有限公司 filed Critical 联咏科技股份有限公司
Priority to CN2010101700703A priority Critical patent/CN102236885A/en
Publication of CN102236885A publication Critical patent/CN102236885A/en

Links

Abstract

The invention relates to a filter for reducing image noise and a filtering method. The filter for reducing image noise comprises an absolute difference total sum calculating unit and a weight affording unit. The absolute difference total sum calculating unit is used for receiving multiple pixels of a target window area and receiving multiple pixels of multiple surrounding window areas corresponding to the surrounding of a target pixel of the target window area. Each surrounding window area has a surrounding pixel around the target pixel. The calculating unit is used for counting a difference absolute value between the target window area and each pixel corresponding to the surrounding window areas, and the difference absolute values are processed by difference calculation to obtain a difference analysis value. The weight affording unit is used for receiving each difference analysis value and obtaining multiple weight values according to a data table, wherein the multiple weight values respectively correspond to the surrounding pixels.

Description

减少图像噪声的过滤器与过滤方法 Filter and reduce image noise filtering method

技术领域 FIELD

[0001] 本发明涉及一种图像噪声的过滤技术,在减少图像噪声时,图像中的细节仍可以有相当程度的保留。 [0001] The present invention relates to an image noise filtering technique, while reducing image noise, details in the image can still be a considerable degree of retention.

背景技术 Background technique

[0002] 一张数字图像是由多个像素以阵列方式所组成,每一个像素,以分别显示所要的颜色与灰阶。 [0002] a digital image is composed of a plurality of pixels in a matrix manner, each pixel to display color and grayscale are desired. 就实际的图像,如果像素受到干扰而显示不适当的灰阶值,就会造成图像噪声。 On the actual image, if the pixel inappropriate interference on display grayscale value, it will cause image noise. 因此,图像显示时需要适当的过滤处理,以调整每一个像素的实际显示的灰阶值。 Thus, appropriate filtering the image is displayed, to adjust actual gray level value of each pixel of the display.

[0003] 过滤处理会调整像素的灰阶值以消除噪声。 [0003] The filtering process adjusts the pixel grayscale values ​​to eliminate noise. 然而如果过度调整以消除噪声,则图像的细节也会被减弱,例如导致图像不清晰。 However, if adjusted to eliminate excessive noise, the details of the image will also be reduced, for example, causing unclear images.

[0004] 图1绘示传统图像噪声过滤技术的处理方式示意图。 Schematic [0004] FIG 1 illustrates a conventional image processing mode noise filtering techniques. 参与图1,对于一个目标像素104而言,其与周围相邻的像素构成一过滤窗区(filtering window) 1020当过滤窗区102的像素值与目标像素104的像素值接近时,可以通过参考过滤窗区的像素值,来估计原本目标像素104未受污染前的像素值,以消除目标像素104的噪声成分。 When participating in FIG. 1, for a target pixel 104, its surrounding neighboring pixels form a window region filter (filtering window) 1020 when the filter window area and the pixel value of the target pixel 104 is close to 102, by reference filtering the pixel values ​​of the window region to estimate the original pixel value before the target pixel 104 uncontaminated, to eliminate the noise component of the target pixel 104. 然而,图像上会一有些图像细节,例如是物件的边缘100。 However, some images will be a detail on an image, for example, the edges of objects 100. 如果,过滤窗区102的周围像素106涵盖到物件边缘(edge) 100时,在过滤噪声的同时,边缘100的特性也会被平滑减弱。 If the filter window around the pixel region 102 to the object 106 to cover the edge (Edge) 100, while the filtering of noise, the characteristics of the edge 100 is also smoothed weakened. 如果调整程度太强, 则边缘100特性就明显被过度减弱甚至消失,影响图像品质。 If the degree of adjustment is too strong, then the 100 properties on the edge of being over-obviously weaken or even disappear, affecting image quality.

[0005] 就一般的过滤技术,例如Sigma过滤即一般采用的技术。 [0005] For general filtration techniques, e.g. Sigma filtering technique which is generally used. 图2绘示Sigma过滤技术的流程示意图。 FIG 2 is a schematic flow diagram schematic Sigma filtering technique. 参阅图1与图2,差异计算单元120会接收目标像素104以及与目标像素104周围相邻的周围像素106的灰阶值。 Referring to FIG. 1 and FIG. 2, the difference calculation unit 120 receives the target pixel 104 and surrounding the target pixel 104 and surrounding pixels adjacent to the grayscale value 106. 差异计算单元120计算周围像素106与目标像素104的差异绝对值。 Difference calculating unit 120 calculates the difference of the target pixel 106 and surrounding pixels of the absolute value of 104. 接着权重计算单元122,根据通过每一个周围像素106分别的差异绝对值,经查表方式取得每一个周围像素106的权重值。 Then weight calculation unit 122, the absolute value of the surrounding pixels 106 through each respective differences around each pixel to obtain a weight value by the lookup table 106. 此权重值可以在过滤窗区102做平均, 以调整目标像素104的灰阶值。 This weight value may be averaged in the filter window region 102 to adjust the gray level value of the target pixel 104.

[0006] 上述的传统过滤方式有可能会对图像细节过度调整而失去图像细节的锐利度。 [0006] The conventional filtering image details might have excessive loss of image sharpness adjustment details. 发明内容 SUMMARY

[0007] 本发明提供一种减少图像噪声的过滤技术,至少可以在过滤图像噪声的同时,尽可能地保留图像中图像细节的内容。 [0007] The present invention provides a method of reducing image noise filtering technique, the image may be simultaneously noise filter, retains at least as much as possible the image details of the content image.

[0008] 本发明提供一种减少图像噪声的过滤器,包括一绝对差异总和计算单元与一权重给予单元。 [0008] The present invention provides a method of reducing image noise filter, comprising a sum of absolute difference calculation unit and a unit weight is given. 绝对差异总和计算单元接收一目标窗区的多个像素以及接收相对该目标窗区的一目标像素周围的多个周围窗区的多个像素。 The plurality of pixels a plurality of sum of absolute difference calculation unit receives a target window region, and a plurality of surrounding pixels around the target region of the receiving window relative to the target region of the window. 每一个周围窗区有一个周围像素在目标像素的周围。 Each area has a window around the surrounding pixels around the target pixel. 计算单元计算目标窗区与周围窗区对应的每一个该像素的一差异绝对值,将这些差异绝对值做一差异计算得到一差异分析值。 Calculation unit calculates a difference of each of the pixels of the target zone and the window surrounding the window region corresponding to the absolute value of the absolute values ​​of these differences to make a difference between a calculated value of the difference analysis. 权重给予单元接收每一个该差异分析值,根据一数据表得到多个权重值,分别对应这些周围像素。 Weight is given to each unit receives the difference value analysis to obtain a plurality of weight values ​​according to the data table, the corresponding peripheral pixels.

[0009] 本发明提供一种减少图像噪声的过滤方法,用于对一图像做噪声过滤。 [0009] The present invention provides a method of reducing image noise filtering for noise filtering of an image do. 此方法包括针于一目标像素决定一目标窗区,该目标窗区具有一像素图案。 This method comprises determining a target window needle region in a target pixel, the target area having a pixel pattern window. 此方法又包括以该目标像素为参考决定多个周围像素。 This method also includes a reference to the target pixel is determined a plurality of surrounding pixels. 针对每一个该周围像素决定一周围窗区,其中周围窗区也具有该像素图案。 Determining a window around a region for each of the surrounding pixels, wherein a window around the pixel region have a pattern. 计算目标窗区与周围窗区对应的每一个该像素的一差异绝对值。 A difference calculation of each pixel of the target region and surrounding the window region corresponding to the absolute value of the window. 将这些差异绝对值做一差异计算得到一差异分析值。 These differences make the absolute value of a difference value calculated by a difference analysis. 根据分别的每一个该差异分析值,经查表给出多个权重值,分别对应这些周围像素。 The analysis of each of the difference values, the look-up table is given by the plurality of weight values, respectively corresponding to the surrounding pixels.

[0010] 为让本发明的上述特征和优点能更明显易懂,下文特举实施例,并配合附图作详细说明如下。 [0010] In order to make the above features and advantages of the invention more comprehensible, several embodiments accompanied with figures are described in detail below.

附图说明 BRIEF DESCRIPTION

[0011] 图1绘示传统图像噪声过滤技术的处理方式示意图。 Schematic [0011] FIG 1 illustrates a conventional image processing mode noise filtering techniques.

[0012] 图2绘示Sigma过滤技术的流程示意图。 [0012] FIG. 2 shows a schematic flow filtration technology Sigma.

[0013] 图3绘示依据本发明一实施例,减少图像噪声的过滤器的操作机制示意图。 [0013] FIG. 3 shows a schematic view of the operating mechanism according to an embodiment of the present invention, the filter reduces image noise.

[0014] 图4绘示依据本发明一实施例,周围窗区的示意图。 [0014] FIG. 4 is a schematic view illustrating a region around the window embodiment of the invention, shown basis.

[0015] 图5绘示依据本发明一实施例,目标窗区的示意图。 [0015] FIG. 5 is a schematic diagram illustrating the target area to an embodiment of the window of the present invention, is shown based.

[0016] 图6绘示依据本发明一实施例,取得SAD权重的操作机制示意图。 [0016] FIG. 6 shows a schematic view of the operating mechanism according to a embodiment of the present invention, to obtain SAD weight.

[0017] 图7-图9绘示依据本发明一实施例,SAD窗区的形状选择。 [0017] Figures 7-9 illustrates examples of the shape of a selected region of the window SAD embodiment of the present invention.

[0018]【主要元件符号说明】 [0018] The main reference numerals DESCRIPTION

[0019] 100:边缘 [0019] 100: edge

[0020] 102 :过滤窗区 [0020] 102: Filter WINDOW

[0021] 104:目标像素 [0021] 104: target pixel

[0022] 106:周围像素 [0022] 106: surrounding pixels

[0023] 120 :差异计算单元 [0023] 120: difference calculation means

[0024] 122 :权重计算单元 [0024] 122: weight calculation unit

[0025] 130 =SAD 计算单元 [0025] 130 = SAD calculation unit

[0026] 132 :权重给予单元 [0026] 132: weight to the unit

[0027] 200 =SAD 计算单元 [0027] 200 = SAD calculation unit

[0028] 202 :权重给予单元 [0028] 202: weight to the unit

[0029] 210,220,230 =SAD 窗区 [0029] 210,220,230 = SAD WINDOW

具体实施方式 Detailed ways

[0030] 本发明至少考虑消除图像噪声的同时又能够尽可能保留图像细节。 [0030] The present invention contemplates elimination of at least the image noise while preserving image detail can be possible. 本发明提出减少图像噪声的过滤技术。 The present invention proposes to reduce image noise filtering techniques. 以下举一些实施例来说明本发明,但是本发明不仅限于所举的实施例。 For some embodiments of the present invention will be described, but the present invention is not limited to the embodiments cited. 又,所举的实施例之间可相互结合。 Furthermore, embodiments may be combined with each other embodiments cited.

[0031] 图3绘示依据本发明一实施例,减少图像噪声的过滤器的操作机制示意图。 [0031] FIG. 3 shows a schematic view of the operating mechanism according to an embodiment of the present invention, the filter reduces image noise. 参阅图3,本发明提出绝对差异总和(Sum of Absolute Difference, SAD)计算单元130,对目标像素与周围像素之间做差异分析。 Referring to Figure 3, the present invention provides a sum of absolute difference (Sum of Absolute Difference, SAD) calculation unit 130, between the target pixel and the surrounding pixels do difference analysis. SAD计算单元130接收一目标窗区的多个像素以及接收相对该目标窗区的一目标像素周围的多个周围窗区的多个像素,每一个该周围窗区有一个周围像素在该目标像素的周围。 A plurality of pixel SAD calculation unit receives a plurality of target window region 130 and a plurality of pixels around the pixel of a region around the target window of the target window relative to the receiving area, each of the peripheral area has a window around a pixel in the target pixel around.

[0032] 在描述SAD计算单元130的计算方式前先描述目标窗区与周围窗区的定义。 [0032] The first description defines a target zone and the window area of ​​the window around the previously described calculation SAD calculation unit 130. 图4 绘示依据本发明一实施例,周围窗区的示意图。 FIG 4 shows a schematic diagram according to embodiment, a region around the window embodiment of the present invention. 图5绘示依据本发明一实施例,目标窗区的示意图。 FIG. 5 shows a schematic diagram according to an embodiment of the present invention, the target area of ​​the window shown. 参阅图4-图5,目标窗区以C表示例如是由像素Ctl-C6W 7个像素所构成。 See FIGS. 4-5, the target area is represented by C, for example, the window is constituted by the pixel Ctl-C6W 7 pixels. 周围窗区以N表示例如是由像素Ntl-N6的7个像素所构成。 N represents a window around the area to be constituted by, for example, seven pixels of Ntl-N6. 窗区的形状取决于像素阵列的排列方式以及所选择的形状,也就是像素图案(Pixel pattern)的形状。 The shape of the window depends on the arrangement of the pixel region and the shape of the selected array, i.e. the pattern shape of the pixel (Pixel pattern) of. 一个目标窗区会有一个目标像素Q。 A target window area will have a target pixel Q. 一个周围窗区会有一个周围像素队。 A window around the area surrounding pixels have a team. 周围像素Ntl是指相对于目标像素Ctl 的周围像素。 Ntl refers to peripheral pixels with respect to surrounding pixels of the target pixel Ctl. 在本实施例周围像素Ntl例如是选取与目标像素Ctl直接相邻接的6个周围像素。 Ntl periphery of the pixel in the present embodiment, for example, selecting a target pixel Ctl directly adjacent six surrounding pixels. 又,以目标像素Ctl为参考,依照所要的像素图案的形状选择始于目标窗区的6个邻近像素C1-C6组成目标窗区。 And, Ctl target pixel as a reference, in the shape of the desired pattern of selected pixels begins six adjacent pixels of the target window area consisting of C1-C6 target window area. 在相同形状下,也以周围像素Ntl做参考,选取邻近像素N1-N6构成周围窗区。 Under the same shape, but also a reference to surrounding pixels Ntl select neighboring pixels around the N1-N6 constituting the window area. 然而,目标窗区与周围窗区的形状相同,但是形状的选择不必定是如图4-图5的选择方式,其后续于图7-图9会有说明。 However, the shape of the target window area around the window area of ​​the same, but the choice is to select the shape is not necessarily 4- embodiment of FIG. 5 in the figure, which in the subsequent Figures 7-9 will be described.

[0033] 当选定目标窗区与周围窗区的形状,例如本实施例的图4-图5所示,则以窗区为单位计算像素其间的差异,例如是灰阶值的差异,或是也可以是其他特性值需要处理的差异。 [0033] When the selected target area and the shape of the window area around the window, for example, the present embodiment of FIGS. 4 to 5, places the window as the unit pixel difference therebetween is calculated, for example, a difference of grayscale value, or can also be a difference in the value of other properties need to be addressed.

[0034] 回到图3,SAD计算单元130是计算目标窗区与周围窗区对应的每一个该像素的一差异绝对值,将这些差异绝对值做一差异计算得到一差异分析值。 [0034] Returning to FIG 3, SAD calculation unit 130, a difference is calculated for each pixel of the target region and surrounding the window region corresponding to the absolute value of the window, the absolute values ​​of these differences to make a difference between a calculated value of the difference analysis. 在一实施例更详细而言就是先计算分别像素C^.j与像素Ncm,...,6的绝对差异值。 In a more detailed embodiment of the first embodiment is calculated in terms of each pixel and the pixel C ^ .j Ncm, ..., the absolute difference value of 6. 在一实施例,SAD计算单元130会把7个绝对差异值作加总(sum)得到对应此周围像素Ntl的窗区差异值。 This give the corresponding peripheral pixels a difference value of the window region Ntl In one embodiment, SAD calculation unit 130 will 7 for an absolute difference value of one embodiment summing (sum). 周围像素N。 N. surrounding pixels 相对目标像素Ctl有多个。 Ctl have more pixels relative to the target. 依照相同方式分别计算出每一个周围像素Ntl的窗区差异值。 It was calculated difference value of the window around each pixel area according to the same manner as Ntl.

[0035] 又,依照差异分析的方式,绝对差异值也可以先做其他运算,例如先做平方或是其他羃次的计算之后才做加总,又或是也可以依照其他差异分析机制的得出可以反映出差异的差异分析值。 [0035] Also, in accordance with the way of variance analysis, the absolute value of the difference can do first other operations, such as square or other Mili do first after calculation times before doing the sum, or they may also have other differences analysis in accordance with the mechanisms the difference may reflect the difference in value analysis. 又,当目标像素是在实际图像的边界时,窗区的像素可能会超过边界,则超过的相素可以设定为零或是一预定值,以利于计算。 Also, when the target pixel is a boundary of an actual image, the pixel region may exceed the window boundary, the phase factors may be exceeding a predetermined value or set to zero, in order to facilitate calculations.

[0036] 当SAD计算单元130计算出每一个周围像素相对于目标像素的差异分析值后给后续的权重给予单元132,以分别得到周围像素的权重值。 [0036] When the SAD calculation unit 130 calculates around each pixel with respect to the value of the pixel difference analysis to the subsequent unit weight to 132 to respectively the weight values ​​of surrounding pixels. 权重给予单元132例如是根据一数据表得到多个权重值,分别对应这些周围像素。 Weighting unit 132, for example, administered to give a plurality of weight values ​​according to the data table, the corresponding peripheral pixels. 数据表可以是经验所得到的数据,或是开放给使用者自行设定的数个选项。 Data tables can be obtained by empirical data, or open to the users themselves set several options. 换句话说,经由查表方式可以得到要给予周围像素的权重, 供后续目标像素的平均处理,以调整目标像素的强度,例如是灰阶值的调整。 In other words, the right can be obtained via a lookup table to give a heavy surrounding pixels, for subsequent averaging processing target pixel, to adjust the intensity of the target pixel, for example, to adjust the gray level values.

[0037] 像素平均的方式,例如是依权重做平均,其中目标像素也例如可以有其本身的权重值,其取决于所采用的平均方式。 [0037] The average pixel manner, for example by the redo weighted average, where for example, the target pixel may also have its own weighting value, which depends on the mean embodiment employed. 权重值的给予原则一般是差异值愈大则权重值愈小,如此可以保留更多边缘的细节,而平滑(smooth)其他区域的细节,以减少噪声。 Principle weight values ​​given are generally the larger the value of the weight difference value is smaller, so the edge can retain more details, the details of other regions and smooth (Smooth), to reduce noise.

[0038] 根据上述的相同概念,在SAD计算单元130的差异分析也可以同时针对像素以另一个权重方式做差异计算。 [0038] The same concept as described above, the difference analysis unit 130 calculates the SAD may be done simultaneously for a pixel difference calculated in another manner heavy weights. 图6绘示依据本发明一实施例,减少图像噪声的过滤器的操作机制示意图。 FIG. 6 shows a schematic diagram according to one embodiment of the operation mechanism of the present invention, the filter reduces image noise. 参阅图6,相对于SAD计算单元200的目标窗区与周围窗区也是如图4-图5 所述,而SAD计算单元200的差异计算方式也与图3的SAD计算单元相似,而其间差异是在计算像素Ccm,...,6与像素Ncm,...,6的绝对差异值时,又分别给一组权重值对应窗区内的每一个像素差异。 Referring to Figure 6, the SAD calculation target window relative to the region and the surrounding region of the window unit 200 is shown in FIG. 4 to FIG 5, and the difference calculated SAD calculation unit 200 calculates a SAD also FIG. 3 units similar, and differences therebetween when calculating the pixel is Ccm, ..., the pixel Ncm 6, ..., the absolute difference value of 6, respectively, and a set of weights to a weight value of the corresponding pixel of each difference window area. 权重值也是可以根据查表方式或是开放给使用者设定而取得。 Weight value also can be made according to the lookup table or open to user settings.

[0039] 接着,权重给予单元202如图3的权重给予单元132相同,会分别对每一个SAD窗区给一个权重值,以供平均计算的使用。 [0039] Next, the weight is given the right to use the same unit 3 given a weight 132, respectively, will be given a weight value for each window area SAD, for average calculation unit 202 shown in FIG. SAD窗区的权重值是给予SAD窗区的代表像素,例如是目标像素以及相对于目标像素的周围像素。 Weighting value of the SAD is representative pixel window region SAD given window area, for example, with respect to the target pixel and pixels surrounding the target pixel.

[0040] 关于SAD窗区的像素图案的形状,除了图4-图5的实施例外,也可以采用不直接邻接的方式来选取,有其数量也不限制最相邻的周围像素。 [0040] The shape of the pixel pattern of the SAD of the window region, addition to the embodiment of FIGS. 4-5, may be employed not be selected directly adjacent manner, there is not limit the number of most adjacent pixels around. 图7-图9绘示依据本发明一实施例,SAD窗区的形状选择。 Figures 7-9 illustrate examples of the shape of a selected region of the window SAD embodiment of the present invention. 但是图7-图9是用来描述可以有其他变化,但不是仅有的变化选择方式。 However, FIGS. 7-9 is used to describe other possible variations, but not the only way to change selection.

[0041] 参阅图7,以三条像素为例,在紧邻像素的选择方式下,以C像素为目标像素,则SAD窗区210的取样点可以是连续超过一个以上的像素,其总数量也不限于周围的8个像 [0041] Referring to Figure 7, an example to three pixels in the proximate pixel selection method, the target pixel to pixel C, then the SAD sample points the window region 210 may be more than one pixel in a row, the total number is not limited to eight surrounding image

ο ο

[0042] 参阅图8,以C像素为目标像素,以三条像素为例,SAD窗区220的取样点可以是间 [0042] Referring to Figure 8, as the target pixel to pixel C, to three pixels as an example, the sampling point SAD window region 220 may be between

隔一个像素的像素图案。 A pattern of pixels of the pixel partition.

[0043] 参阅图9,以C像素为目标像素,以三条像素为例,SAD窗区230的取样点可以是间 [0043] Referring to FIG. 9, the pixel C as the target pixel to three pixels as an example, the sampling point SAD window region 230 may be between

隔二个像素的像素图案。 Compartment two pixel pattern of pixels.

[0044] 换句或说,SAD窗区的形状可以依照实际做选择,且相同的一张图像内也允许不同区域有不同形状的SAD窗区。 [0044] In other sentences or said shape of the window region SAD may be selected in accordance with practical, within one and the same image is also allowed to have different regions of the window regions of different shapes SAD.

[0045] 本发明提出在图像过滤过程中,以SAD窗区来考虑差异,取代仅考虑分别单一像素的差异。 [0045] The present invention proposes an image filtering process to the window area to take into account differences SAD, substituted consider only differences were a single pixel. 如此,本申请至少在过滤的处理应用上可以保留更多的图像细节。 Thus, the present application at least on the filtering processing application may retain more image detail.

[0046] 虽然本发明已以实施例公开如上,然其并非用以限定本发明,本领域技术人员,在不脱离本发明的精神和范围内,当可作些许的更动与润饰,故本发明的保护范围当视所附权利要求书所界定者为准。 [0046] While the invention has been disclosed in the above embodiments, they are not intended to limit the invention, those skilled in the art, without departing from the spirit and scope of the present invention, may make various modifications and variations, so the depending on the scope of the invention as defined by the appended claims and their equivalents.

Claims (24)

1. 一种减少图像噪声的过滤器,包括:一绝对差异总和计算单元,接收一目标窗区的多个像素以及接收相对该目标窗区的一目标像素周围的多个周围窗区的多个像素,每一个该周围窗区有一个周围像素在该目标像素的周围,其中该绝对差异总和计算单元计算该目标窗区与该周围窗区对应的每一个该像素的一差异绝对值,将这些差异绝对值做一差异计算得到一差异分析值;以及一权重给予单元,接收每一个该差异分析值,根据一数据表得到多个权重值,分别对应这些周围像素。 1. A method of reducing image noise filter, comprising: a plurality of sum of absolute difference calculating unit, receiving a plurality of pixels and a plurality of target window area around a pixel area around the target window of the target window relative to the receiving region pixels, each of which has around the window area around a pixel around the target pixel, wherein the sum of the absolute difference calculating unit calculates a difference of each of the pixels in the target window region and the peripheral region corresponding to the absolute value of the window, these make a difference in the absolute value of the difference calculated by analyzing a difference value; and a weight applying means, each receiving a value of the difference analysis, to obtain a plurality of weight values ​​according to the data table, the corresponding peripheral pixels.
2.如权利要求1所述的减少图像噪声的过滤器,其中该目标窗区具有一像素图案,该目标窗区的位置对应该目标像素,,以及该周围窗区与该目标窗区具有相同的该像素图案, 且该周围窗区的位置对应该周围像素。 2. The filter according to claim 1 to reduce the image noise, wherein the target region having a pixel pattern window, the window position of the target area of ​​the target pixel should ,, and the window region surrounding the target window has the same area the position of the pixel pattern, and the area around the window to be surrounding pixels.
3.如权利要求2所述的减少图像噪声的过滤器,其中在该像素图案内的这些像素是直接相邻。 2, the filter image noise reduction as claimed in claim 3, wherein the pixels in the pixel pattern is directly adjacent.
4.如权利要求2所述的减少图像噪声的过滤器,其中在该像素图案内的这些像素不是全部直接相邻。 4. The filter according to claim 2 reduce the image noise, wherein the pixels in the pixel pattern is not all immediately adjacent.
5.如权利要求1所述的减少图像噪声的过滤器,其中该绝对差异总和计算单元计算该差异分析值是直接取这些差异绝对值做加总所得到。 5. The filter according to claim 1 to reduce the image noise, wherein the sum of the absolute difference calculation unit calculates the difference value is taken directly analyze these differences make the absolute value sum is obtained.
6.如权利要求1所述的减少图像噪声的过滤器,其中该绝对差异总和计算单元计算该差异分析值是直接取这些差异绝对值再乘以一调整权重值后,做加总所得到。 6. The filter according to claim 1 to reduce the image noise, wherein the sum of the absolute difference calculation unit calculates the difference value analysis of these differences are directly taken after the absolute value multiplied by a weight value adjusted weights, do sum obtained.
7.如权利要求6所述的减少图像噪声的过滤器,其中该调整权重值是可调整的。 7. The filter according to claim 6 to reduce image noise, wherein the adjusted weight value is adjustable.
8.如权利要求1所述的减少图像噪声的过滤器,其中该绝对差异总和计算单元计算该差异分析值是取这些差异绝对值的平方后做加总所得到。 8. The filter according to claim 1 to reduce the image noise, wherein the sum of the absolute difference calculation unit calculates the difference value is analyzed after taking the absolute value of these differences make the square sum obtained.
9.如权利要求1所述的减少图像噪声的过滤器,其中该绝对差异总和计算单元计算该差异分析值是取这些差异绝对值的平方再乘以一调整权重值后,做加总所得到。 9. The filter according to claim 1 to reduce the image noise, wherein the sum of the absolute difference calculation unit calculates the difference value analysis is to take the square of the absolute difference value multiplied by a weight to adjust the weights, do sum obtained .
10.如权利要求9所述的减少图像噪声的过滤器,其中该调整权重值是可调整的。 10. The filter according to claim 9 to reduce image noise, wherein the adjusted weight value is adjustable.
11.如权利要求1所述的减少图像噪声的过滤器,其中该目标窗区与该周围窗区具有相同的像素图案,且该像素图案是一固定图案。 11. The filter according to claim 1 to reduce the image noise, wherein the target window region and the peripheral region have the same pixel window pattern and the pixel pattern is a fixed pattern.
12.如权利要求1所述的减少图像噪声的过滤器,其中该目标窗区与该周围窗区具有相同的像素图案,该像素图案会依照图像内容变化。 12. The filter according to claim 1 to reduce the image noise, wherein the target window region and the region surrounding the window has the same pixel pattern, pixel pattern vary in accordance with the image content.
13. 一种减少图像噪声的过滤方法,用于对一图像做噪声过滤,包括:针于一目标像素决定一目标窗区,该目标窗区具有一像素图案;以该目标像素为参考决定多个周围像素;针对每一个该周围像素决定一周围窗区,该周围窗区也具有该像素图案;计算该目标窗区与该周围窗区对应的每一个该像素的一差异绝对值;将这些差异绝对值做一差异计算得到一差异分析值;以及根据分别的每一个该差异分析值,经查表给出多个权重值,分别对应这些周围像素。 13. A method of reducing image noise filtering, for images having a noise filtering, comprising: determining a target window of a needle to a target pixel area, the target area having a pixel pattern window; reference to the target pixel determines how a surrounding pixels; for each of the surrounding pixels, determining a region around the window, the window around the pixel region also having a pattern; a difference of the pixel is calculated for each window of the target region and the peripheral region corresponding to the absolute value of the window; these make a difference in the absolute value of the difference calculated by analyzing a difference value; and analysis according to each of the difference values, the look-up table is given by the plurality of weight values, respectively corresponding to the surrounding pixels.
14.如权利要求13所述的减少图像噪声的过滤方法,其中该目标窗区的位置对应该目标像素,以及该周围窗区的位置对应该周围像素。 14. The filtering method of claim 13 wherein reduction of image noise of the target position of the window area should be the target pixel position, and the peripheral area of ​​the window to be surrounding pixels claim.
15.如权利要求14所述的减少图像噪声的过滤方法,其中在该像素图案内的这些像素是直接相邻。 15. The method of filtering image noise reduction according to claim 14, wherein the pixels in the pixel pattern is directly adjacent.
16.如权利要求14所述的减少图像噪声的过滤方法,其中在该像素图案内的这些像素不是全部直接相邻。 16. The method of filtering image noise reduction according to claim 14, wherein the pixels in the pixel pattern is not all immediately adjacent.
17.如权利要求13所述的减少图像噪声的过滤方法,其中计算该差异分析值是直接取这些差异绝对值做加总所得到。 17. The method of filtering image noise reduction according to claim 13, wherein calculating the difference value is taken directly analysis of these absolute difference sum obtained do.
18.如权利要求13所述的减少图像噪声的过滤方法,其中计算该差异分析值是直接取这些差异绝对值再乘以一调整权重值后,做加总所得到。 18. After the filtration method according to claim 13 to reduce the image noise, wherein calculating the difference value is taken directly analysis of these absolute difference value multiplied by a weight to adjust the weights do the resulting sum.
19.如权利要求18所述的减少图像噪声的过滤方法,还包括调整该调整权重值。 19. The method of filtering image noise reduction according to claim 18, further comprising adjusting a weight value of the weight adjustment.
20.如权利要求13所述的减少图像噪声的过滤方法,其中计算该差异分析值是取这些差异绝对值的平方后做加总所得到。 20. The method of filtering image noise reduction according to claim 13, wherein calculating the difference value is analyzed after taking the absolute value of these differences make the square sum obtained.
21.如权利要求13所述的减少图像噪声的过滤方法,其中该绝对差异总和计算单元计算该差异分析值是取这些差异绝对值的平方再乘以一调整权重值后,做加总所得到。 21. After the filtration method according to claim 13 to reduce the image noise, wherein the sum of the absolute difference calculation unit calculates the difference value analysis is to take the square of the absolute difference value multiplied by a weight to adjust the weights, do sum obtained .
22.如权利要求21所述的减少图像噪声的过滤方法,还包括调整该调整权重值。 22. The method of filtering image noise reduction according to claim 21, further comprising adjusting a weight value of the weight adjustment.
23.如权利要求13所述的减少图像噪声的过滤方法,还包括设定该像素图案为一固定图案。 23. The method of filtering image noise reduction according to claim 13, further comprising a set of the pixel pattern is a fixed pattern.
24.如权利要求13所述的减少图像噪声的过滤方法,还包括设定该像素图案,使该像素图案依照图像内容而具有变化。 24. The method of filtering image noise reduction according to claim 13, further including setting the pixel pattern, so that the pixel pattern in accordance with image content change.
CN2010101700703A 2010-04-21 2010-04-21 Filter for reducing image noise and filtering method CN102236885A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2010101700703A CN102236885A (en) 2010-04-21 2010-04-21 Filter for reducing image noise and filtering method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2010101700703A CN102236885A (en) 2010-04-21 2010-04-21 Filter for reducing image noise and filtering method

Publications (1)

Publication Number Publication Date
CN102236885A true CN102236885A (en) 2011-11-09

Family

ID=44887512

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2010101700703A CN102236885A (en) 2010-04-21 2010-04-21 Filter for reducing image noise and filtering method

Country Status (1)

Country Link
CN (1) CN102236885A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104036457A (en) * 2013-09-02 2014-09-10 上海联影医疗科技有限公司 Image noise reduction method
CN104168405A (en) * 2013-05-20 2014-11-26 聚晶半导体股份有限公司 Noise reduction method and image processing device
CN104685539A (en) * 2012-09-28 2015-06-03 株式会社岛津制作所 Digital image processing method and an imaging apparatus
CN104735300A (en) * 2015-03-31 2015-06-24 中国科学院自动化研究所 Video denoising device and method based on weight filtering

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2002099753A1 (en) * 2001-06-05 2002-12-12 Sony Corporation Image processor
CN1450811A (en) * 2002-04-05 2003-10-22 精工爱普生株式会社 Self-adaptive post-filtering for reducing noise in highly compressed image/video coding
CN1469635A (en) * 2002-06-05 2004-01-21 佳能株式会社 Image processing apparatus, image processing method and computer program
CN101309407A (en) * 2007-05-14 2008-11-19 奇景光电股份有限公司 Movement estimating method
CN101473656A (en) * 2006-06-29 2009-07-01 汤姆森许可贸易公司 Adaptive filtering based on pixel

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2002099753A1 (en) * 2001-06-05 2002-12-12 Sony Corporation Image processor
CN1450811A (en) * 2002-04-05 2003-10-22 精工爱普生株式会社 Self-adaptive post-filtering for reducing noise in highly compressed image/video coding
CN1469635A (en) * 2002-06-05 2004-01-21 佳能株式会社 Image processing apparatus, image processing method and computer program
CN101473656A (en) * 2006-06-29 2009-07-01 汤姆森许可贸易公司 Adaptive filtering based on pixel
CN101309407A (en) * 2007-05-14 2008-11-19 奇景光电股份有限公司 Movement estimating method

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104685539B (en) * 2012-09-28 2018-05-04 株式会社岛津制作所 Digital image processing method and an imaging apparatus
CN104685539A (en) * 2012-09-28 2015-06-03 株式会社岛津制作所 Digital image processing method and an imaging apparatus
CN104168405B (en) * 2013-05-20 2017-09-01 聚晶半导体股份有限公司 Noise suppression method and image processing apparatus
CN104168405A (en) * 2013-05-20 2014-11-26 聚晶半导体股份有限公司 Noise reduction method and image processing device
CN104036457A (en) * 2013-09-02 2014-09-10 上海联影医疗科技有限公司 Image noise reduction method
CN104036457B (en) * 2013-09-02 2015-03-18 上海联影医疗科技有限公司 Image noise reduction method
CN104735300A (en) * 2015-03-31 2015-06-24 中国科学院自动化研究所 Video denoising device and method based on weight filtering
CN104735300B (en) * 2015-03-31 2017-12-01 中国科学院自动化研究所 Video apparatus and method for denoising weight Filtering

Similar Documents

Publication Publication Date Title
US8098953B2 (en) Image processing apparatus and image processing method for adaptively processing an image using an enhanced image and edge data
JP3009726B2 (en) Conversion apparatus and method for an image digital signal
US7636496B2 (en) Histogram adjustment for high dynamic range image mapping
EP1289310A2 (en) Method and system for adaptive demosaicing
KR101128661B1 (en) Image processing apparatus, image processing method, and recording medium in which program of image processing method has been recorded
US20070229863A1 (en) Tone Correction Apparatus, Mobile Terminal, Image Capturing Apparatus, Mobile Phone, Tone Correction Method and Program
JP4577621B2 (en) Image correction processing system and the image correction processing method
US20030161549A1 (en) Contrast enhancement of digital images
JP4804660B2 (en) Noise reduction method using the color information for the digital image processing
JP4707830B2 (en) Method for improving a digital image having a noise dependent control of texture
US9338349B2 (en) Generation of ghost-free high dynamic range images
JP5495025B2 (en) An image processing apparatus and method, and program
JP2003526272A (en) System and method for improving the sharpness of the video image
JP2008159071A (en) Visual processing device, visual processing method, program, display device, and integrated circuit
WO2009089847A1 (en) Image processing method and apparatus
CN102231264B (en) Dynamic contrast enhancement device and method
CN101783012B (en) Automatic image defogging method based on dark primary colour
JPH11150669A (en) Image improving device
JP2007110576A (en) Color correction device for subject image data and control method thereof
CN101527034A (en) Method and device for correlating adaptive frames
JP2012096083A (en) Image processing method and device for detection and emphasis of edge
JP4648071B2 (en) Color saturation control method of the image display apparatus and a video signal
CN102158653A (en) Device and method for acquiring digital image with high dynamic range in real time
JP3731577B2 (en) Image processing program
Zolliker et al. Retaining local image information in gamut mapping algorithms

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
C12 Rejection of a patent application after its publication