JPH0512435A - Noise removing method for image - Google Patents
Noise removing method for imageInfo
- Publication number
- JPH0512435A JPH0512435A JP3158975A JP15897591A JPH0512435A JP H0512435 A JPH0512435 A JP H0512435A JP 3158975 A JP3158975 A JP 3158975A JP 15897591 A JP15897591 A JP 15897591A JP H0512435 A JPH0512435 A JP H0512435A
- Authority
- JP
- Japan
- Prior art keywords
- pixel
- range
- pixels
- pixel value
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Landscapes
- Image Processing (AREA)
- Picture Signal Circuits (AREA)
Abstract
Description
【０００１】[0001]
【産業上の利用分野】この発明は、画像のノイズ除去方
法に関し、特にデジタル信号処理を行うテレビジョン受
像機における画像のノイズ除去方法に関する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image noise eliminating method, and more particularly to an image noise eliminating method in a television receiver for digital signal processing.
【０００２】[0002]
【従来の技術】この種の従来のノイズ除去方法の１つと
して、Ｓｉｇｍａフィルタによる方法が知られている。
このＳｉｇｍａフィルタによる方法では、フィールド内
に１つの注目画素を含むかつ水平方向画素数がｎ0 、垂
直方向画素数がｍ0 の領域（ｎ 0 ×ｍ0 ）を設定し、注
目画素の画素値（Ａ0 ）を中心とする範囲〔Ａ0 −
ｄ0 ，Ａ0 ＋ｄ0 〕を設定し、領域（ｎ0 ×ｍ0 ）内の
画素のうちその画素値が範囲〔Ａ0 −ｄ0 ，Ａ0 ＋
ｄ0 〕に入るすべての画素の平均画素値（Ａ0 ´）を求
め、その平均画素値（Ａ0 ´）を注目画素の元の画素値
（Ａ0 ）と置き換えるようにしている。2. Description of the Related Art One of conventional noise removing methods of this type
Then, a method using a Sigma filter is known.
In this Sigma filter method,
Contains one target pixel and the number of horizontal pixels is n0, Hanging
The number of pixels in the direct direction is m0Area (n 0× m0) And note
Pixel value of eye pixel (A0) Centered around [A0−
d0, A0+ D0], The area (n0× m0)Inside
The pixel value of the pixel is in the range [A0-D0, A0+
d0] The average pixel value (A0´)
Therefore, the average pixel value (A0′) Is the original pixel value of the pixel of interest
(A0).
【０００３】[0003]
【発明が解決しようとする課題】上述のＳｉｇｍａフィ
ルタによる方法では、領域（ｎ0 ×ｍ0 ）内の画素のう
ちその画素値が範囲〔Ａ0 −ｄ0 ，Ａ0 ＋ｄ0 〕に入る
画素が注目画素を除いて１つ以上存在する場合に、注目
画素の画素値をノイズとして平滑化することができる。
したがって、画像全体に均一なホワイトノイズで、その
平均的な信号強度を表す標準偏差が推定可能な場合には
有効であった。In the method using the Sigma filter described above, a pixel whose pixel value falls within the range [A 0 -d 0 , A 0 + d 0 ] among pixels in the region (n 0 × m 0 ). When there are one or more pixels except the pixel of interest, the pixel value of the pixel of interest can be smoothed as noise.
Therefore, it was effective when the white noise was uniform over the entire image and the standard deviation representing the average signal intensity could be estimated.
【０００４】しかし、領域（ｎ0 ×ｍ0 ）内の画素のう
ちその画素値が範囲〔Ａ0 −ｄ0 ，Ａ0 ＋ｄ0 〕に入る
画素が注目画素を除いて１つも存在しない場合には、注
目画素が孤立したノイズとしてそのまま残ってしまうと
いう問題点があった。それゆえに、この発明の主たる目
的は、孤立したノイズを除去することができる、画像の
ノイズ除去方法を提供することである。However, when there is no pixel except the pixel of interest, the pixel value of which falls within the range [A 0 -d 0 , A 0 + d 0 ] among the pixels in the area (n 0 × m 0 ). Has a problem that the pixel of interest remains as isolated noise. Therefore, a main object of the present invention is to provide a method for denoising an image, which can remove isolated noise.
【０００５】[0005]
【課題を解決するための手段】この発明は、フィールド
内に１つの注目画素を設定し、注目画素に隣接する複数
の画素のうちその画素値が注目画素の画素値に近い第１
および第２の画素を求め、注目画素ならびに第１および
第２の画素のそれぞれの画素値をすべて含む最も狭い第
１の範囲を求め、第１の範囲を含むかつそれよりも広い
変更可能な第２の範囲を求め、注目画素および複数の画
素を含む所定領域内の画素のうちその画素値が第２の範
囲に入るすべての画素の平均画素値を求め、平均画素値
を注目画素の元の画素値と置き換える、画像のノイズ除
去方法である。According to the present invention, one target pixel is set in a field, and the pixel value of a plurality of pixels adjacent to the target pixel is close to the pixel value of the target pixel.
And a second pixel are obtained, a narrowest first range including all the pixel values of the pixel of interest and the first and second pixels is obtained, and a changeable first range including the first range and wider than the first range is obtained. 2 is obtained, the average pixel value of all the pixels in the predetermined area including the target pixel and the plurality of pixels whose pixel value falls within the second range is calculated, and the average pixel value is set to the original pixel value of the target pixel. This is a method of removing noise from an image, which replaces pixel values.
【０００６】[0006]
【作用】フィールド内に１つの注目画素を設定し、注目
画素に隣接する複数（たとえば８つ）の周囲画素のうち
その画素値が注目画素の画素値（Ａ）に近い第１および
第２の画素を求め、注目画素ならびに第１および第２の
画素のそれぞれの画素値をすべて含む最も狭い第１の範
囲〔Ｖ1 ，Ｖ2 〕を求め、第１の範囲〔Ｖ1 ，Ｖ 2 〕を
含むかつそれよりも広い変更可能な第２の範囲〔Ｖ1 −
ｄ，Ｖ2 ＋ｄ〕を求め、注目画素およびそれに隣接する
複数の周囲画素を含む所定領域（ｎ×ｍ）内の画素のう
ちその画素値が第２の範囲〔Ｖ1 −ｄ，Ｖ2 ＋ｄ〕に入
るすべての画素の平均画素値（Ａ´）を求め、この平均
画素値（Ａ´）を注目画素の元の画素値（Ａ）と置き換
えて新画素値とする。このように、注目画素ならびに第
１および第２画素を含む少なくとも３つの画素の平均画
素値（Ａ´）を新画素値として用いるので、注目画素の
元の画素値（Ａ）が周囲画素の画素値と大きく違ってい
てもその画素値（Ａ）は、他の少なくとも２つの画素値
によって平滑化され得る。[Operation] One target pixel is set in the field
Of the multiple (for example, eight) surrounding pixels that are adjacent to the pixel
The pixel value whose pixel value is close to the pixel value (A) of the pixel of interest
The second pixel is obtained, and the pixel of interest and the first and second pixels
The narrowest first range containing all the pixel values of each pixel
Box [V1, V2], The first range [V1, V 2]
A changeable second range [V that includes and is wider than that.1−
d, V2+ D] is obtained, and the pixel of interest and its adjacent
A pixel within a predetermined area (n × m) including a plurality of surrounding pixels
The pixel value is in the second range [V1-D, V2+ D]
The average pixel value (A ') of all the pixels
Replace the pixel value (A ') with the original pixel value (A) of the pixel of interest
The new pixel value. Thus, the pixel of interest and the
Average image of at least 3 pixels including 1 and 2 pixels
Since the prime value (A ') is used as the new pixel value,
The original pixel value (A) differs greatly from the pixel values of the surrounding pixels.
However, the pixel value (A) is at least two other pixel values
Can be smoothed by
【０００７】[0007]
【発明の効果】この発明によれば、周囲から孤立したノ
イズを平滑化し、除去することができる。また、注目画
素がエッジ部分にある場合には、第１および第２の画素
が、その画素値が注目画素の画素値に近い線図部分の画
素から抽出されるので、第２の範囲〔Ｖ1 −ｄ，Ｖ2 ＋
ｄ〕には、主として線図部分の画素が含まれることにな
り、求められる新画素値（Ａ´）は元の画素値（Ａ）に
非常に近いものとなる。したがって、エッジ部分は確実
に保存され得る。According to the present invention, noise isolated from the surroundings can be smoothed and removed. In addition, when the pixel of interest is at the edge portion, the first and second pixels are extracted from the pixel of the diagram portion whose pixel value is close to the pixel value of the pixel of interest, so the second range [V 1- d, V 2 +
d] mainly includes pixels in the diagram portion, and the obtained new pixel value (A ′) is very close to the original pixel value (A). Therefore, the edge portion can be surely preserved.
【０００８】この発明の上述の目的，その他の目的，特
徴および利点は、図面を参照して行う以下の実施例の詳
細な説明から一層明らかとなろう。The above-mentioned objects, other objects, features and advantages of the present invention will become more apparent from the detailed description of the embodiments below with reference to the drawings.
【０００９】[0009]
【実施例】この発明の一実施例を図１のフロー図に基づ
いて説明する。ラスタスキャンされるテレビジョン画像
の１つのフィールド内の全画素について、未処理画素が
ある場合（Ｓ１）、ステップＳ２において、図２に示す
ようにフィールド内に１つの注目画素１０を設定し、注
目画素１０に隣接する８つの周囲画素１２のうちその画
素値が注目画素１０の画素値（Ａ）に近い第１画素１２
１および第２画素１２２を抽出する。DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described with reference to the flow chart of FIG. When there is an unprocessed pixel for all pixels in one field of the raster-scanned television image (S1), one target pixel 10 is set in the field as shown in FIG. Of the eight surrounding pixels 12 adjacent to the pixel 10, the first pixel 12 whose pixel value is close to the pixel value (A) of the target pixel 10
The first and second pixels 122 are extracted.
【００１０】次いで、ステップＳ３において、図３から
よくわかるように、注目画素１０ならびに第１画素１２
１および第２画素１２２のそれぞれの画素値を含む最も
狭い第１の範囲〔Ｖ1 ，Ｖ2 〕を求め、さらに、第１の
範囲〔Ｖ1 ，Ｖ2 〕を含むかつそれよりも広い第２の範
囲〔Ｖ1 −ｄ，Ｖ2 ＋ｄ〕を求める。なお、ｄの値は、
ノイズの程度等に応じて適宜変更され得る。Next, in step S3, as can be seen clearly from FIG. 3, the pixel of interest 10 and the first pixel 12 are selected.
The narrowest first range [V 1 , V 2 ] including the pixel values of the first and second pixels 122 is obtained, and the first range [V 1 , V 2 ] is included and wider than that. The range [V 1 −d, V 2 + d] of 2 is obtained. The value of d is
It can be appropriately changed according to the degree of noise and the like.
【００１１】そして、ステップＳ４において、注目画素
１０および８つの周囲画素１２を含むかつ水平方向画素
数がｎ（ｎ≧３）、垂直方向画素数がｍ（ｍ≧３）のマ
スク領域１４（図２）内の画素のうちその画素値が第２
の範囲〔Ｖ1−ｄ，Ｖ2 ＋ｄ〕に入るすべての画素の平
均画素値（Ａ´）を算出し、ステップＳ５において、そ
の平均画素値（Ａ´）を注目画素１０の元の画素値
（Ａ）と置き換える。なお、マスク領域１４の範囲を決
定する値ｎおよびｍは、ノイズの程度等に応じて適宜変
更され得る。Then, in step S4, the mask area 14 including the target pixel 10 and eight surrounding pixels 12 and having the number of horizontal pixels n (n ≧ 3) and the number of vertical pixels m (m ≧ 3) (see FIG. The pixel value of the pixels in 2) is the second
The average pixel value (A ′) of all the pixels within the range [V 1 −d, V 2 + d] is calculated, and in step S5, the average pixel value (A ′) is set to the original pixel value of the target pixel 10. Replace with (A). The values n and m that determine the range of the mask region 14 can be appropriately changed depending on the degree of noise or the like.
【００１２】このようなノイズ除去処理はステップＳ１
においてフィールド内の全画素について処理したことが
確認されるまで繰り返し実行する。この実施例によれ
ば、注目画素１０の新画素値（Ａ´）が、第２範囲〔Ｖ
1 −ｄ，Ｖ2 ＋ｄ〕に入る注目画素１０ならびに第１画
素１２１および第２画素１２２のそれぞれの画素値を含
む少なくとも３つの画素値を平均することによって求め
られるので、図４に示すような周囲から孤立したノイズ
でも確実に平滑化され得る。Such noise removal processing is performed in step S1.
The process is repeated until it is confirmed that all pixels in the field have been processed. According to this embodiment, the new pixel value (A ′) of the target pixel 10 is in the second range [V
1− d, V 2 + d], the target pixel 10 and at least three pixel values including the respective pixel values of the first pixel 121 and the second pixel 122 are obtained by averaging, and therefore, as shown in FIG. Even noise isolated from the surroundings can be surely smoothed.
【００１３】また、注目画素１０がエッジ部分にある場
合には、図５に示すように、第１画素１２１および第２
画素１２２が、その画素値が注目画素１０の画素値
（Ａ）に近い線図部分１６の画素から抽出されるので、
第２の範囲〔Ｖ1−ｄ，Ｖ2 ＋ｄ〕には、主として線図
部分１６の画素が含まれることになる。したがって、注
目画素１０の新画素値（Ａ´）は、元の画素値（Ａ）に
非常に近いものとなり、エッジは確実に保存され得る。When the pixel of interest 10 is at the edge portion, as shown in FIG. 5, the first pixel 121 and the second pixel 121
Since the pixel 122 is extracted from the pixel of the diagram portion 16 whose pixel value is close to the pixel value (A) of the target pixel 10,
The second range [V 1 −d, V 2 + d] mainly includes the pixels of the diagram portion 16. Therefore, the new pixel value (A ′) of the target pixel 10 becomes very close to the original pixel value (A), and the edge can be surely preserved.
【００１４】このような原理に基づくノイズ除去装置１
８の機能的回路構成を図６のブロック図に基づいて説明
する。なお、上述した原理的説明と重複する説明は省略
する。図６においては、ラスタスキャンされた画像信号
が、２画素抽出部２０に与えられるとともに、平均値計
算部２２にその一方入力として与えられる。２画素抽出
部２０では、第１画素１２１および第２画素１２２のそ
れぞれの画素値が求められ、これら２つの画素値および
注目画素１０の画素値が、範囲算出部２４に与えられ
る。範囲算出部２４では、与えられた３つの画素値から
第１の範囲〔Ｖ1 ，Ｖ2 〕が求められる。また、範囲算
出部２４には、数値入力部２６から変更可能な数値ｄが
与えられ、この数値ｄと第１の範囲〔Ｖ1 ，Ｖ2 〕とに
よって第２の範囲〔Ｖ1 −ｄ，Ｖ2 ＋ｄ〕が求められる
（図３）。第２の範囲〔Ｖ1 −ｄ，Ｖ2 ＋ｄ〕は、平均
値計算部２２にその他方入力として与えられる。平均値
計算部２２では、マスク領域１４（図２）内においてそ
の画素値が第２の範囲〔Ｖ1 −ｄ，Ｖ2 ＋ｄ〕に入るす
べての画素の平均画素値（Ａ´）が求められる。そし
て、この平均画素値（Ａ´）が、注目画素１０の新画素
値として出力される。A noise removing device 1 based on such a principle.
The functional circuit configuration of No. 8 will be described based on the block diagram of FIG. It should be noted that the description overlapping with the above-described principle description is omitted. In FIG. 6, the raster-scanned image signal is supplied to the two-pixel extraction unit 20 and the average value calculation unit 22 as one input. In the 2-pixel extraction unit 20, the pixel values of the first pixel 121 and the second pixel 122 are obtained, and these two pixel values and the pixel value of the target pixel 10 are given to the range calculation unit 24. The range calculator 24 obtains the first range [V 1 , V 2 ] from the given three pixel values. Further, the range calculation unit 24 is given a changeable numerical value d from the numerical value input unit 26, and the numerical value d and the first range [V 1 , V 2 ] are used to determine the second range [V 1 -d, V 2 + d] is required (FIG. 3). The second range [V 1 −d, V 2 + d] is given to the average value calculation unit 22 as the other input. The average value calculation unit 22 obtains the average pixel value (A ′) of all the pixels in the mask area 14 (FIG. 2) whose pixel values fall within the second range [V 1 −d, V 2 + d]. . Then, this average pixel value (A ′) is output as the new pixel value of the target pixel 10.
【図１】この発明の一実施例を示すフロー図である。FIG. 1 is a flow chart showing an embodiment of the present invention.
【図２】注目画素とマスク領域を示す図解図である。FIG. 2 is an illustrative view showing a target pixel and a mask area.
【図３】注目画素ならびに第１および第２画素のそれぞ
れの画素値と第１および第２の範囲との関係を示す図解
図である。FIG. 3 is an illustrative view showing a relationship between a pixel value of a pixel of interest and first and second pixels and first and second ranges.
【図４】注目画素が周囲から孤立したノイズ部分である
場合の各画素の画素値と第１および第２の範囲との関係
を示す図解図である。FIG. 4 is an illustrative view showing a relationship between a pixel value of each pixel and a first range and a second range when a target pixel is a noise portion isolated from the surroundings.
【図５】注目画素がエッジ部分にある場合の各画素の画
素値と第１および第２の範囲との関係を示す図解図であ
る。FIG. 5 is an illustrative view showing a relationship between a pixel value of each pixel and a first range and a second range when a target pixel is at an edge portion.
【図６】この発明の一実施例を示す機能ブロック図であ
る。FIG. 6 is a functional block diagram showing an embodiment of the present invention.
１０ …注目画素 １２ …周囲画素 １４ …マスク領域 １８ …ノイズ除去装置 ２０ …２画素抽出部 ２２ …平均値計算部 ２４ …範囲算出部 ２６ …数値入力部 10 ... Pixel of interest 12 ... Surrounding pixels 14 ... Mask area 18 ... Noise removal device 20 ... 2 pixel extraction unit 22 ... Average value calculation unit 24 ... Range calculation unit 26 ... Numerical value input unit
Claims (1)
前記注目画素に隣接する複数の画素のうちその画素値が
前記注目画素の画素値に近い第１および第２の画素を求
め、 前記注目画素ならびに前記第１および第２の画素のそれ
ぞれの画素値をすべて含む最も狭い第１の範囲を求め、
前記第１の範囲を含むかつそれよりも広い変更可能な第
２の範囲を求め、 前記注目画素および前記複数の画素を含む所定領域内の
画素のうちその画素値が前記第２の範囲に入るすべての
画素の平均画素値を求め、 前記平均画素値を前記注目画素の元の画素値と置き換え
る、画像のノイズ除去方法。What is claimed is: 1. A target pixel is set in a field,
Of the plurality of pixels adjacent to the target pixel, first and second pixels whose pixel values are close to the pixel value of the target pixel are obtained, and the target pixel and the pixel values of the first and second pixels, respectively. Find the narrowest first range containing all
A changeable second range including the first range and wider than the first range is obtained, and the pixel value of the pixels in the predetermined area including the target pixel and the plurality of pixels falls within the second range. An image noise removing method, wherein an average pixel value of all pixels is obtained, and the average pixel value is replaced with the original pixel value of the pixel of interest.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP3158975A JPH0512435A (en) | 1991-06-28 | 1991-06-28 | Noise removing method for image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP3158975A JPH0512435A (en) | 1991-06-28 | 1991-06-28 | Noise removing method for image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPH0512435A true JPH0512435A (en) | 1993-01-22 |
Family
ID=15683469
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP3158975A Withdrawn JPH0512435A (en) | 1991-06-28 | 1991-06-28 | Noise removing method for image |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPH0512435A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1995016322A3 (en) * | 1993-12-02 | 1995-07-20 | Philips Electronics Nv | Noise reduction |
-
1991
- 1991-06-28 JP JP3158975A patent/JPH0512435A/en not_active Withdrawn
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1995016322A3 (en) * | 1993-12-02 | 1995-07-20 | Philips Electronics Nv | Noise reduction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6907144B1 (en) | Noise reduction method, apparatus, and program for digital image processing | |
| JP2016505186A (en) | Image processor with edge preservation and noise suppression functions | |
| US7110044B2 (en) | Image detail enhancement system | |
| EP0114961B1 (en) | Nonlinear filtering of gray scale video images | |
| US20090041371A1 (en) | Image processing apparatus | |
| CN105096330A (en) | Image processing method capable of automatically recognizing pure-color borders, system and a photographing terminal | |
| WO2016095541A1 (en) | Image processing method, device, system and computer storage medium | |
| WO2004051573A3 (en) | Method and apparatus for removing false edges from a segmented image | |
| US8437572B2 (en) | Method for sharpness enhancing an image | |
| CN111986095B (en) | Image processing method and image processing device based on edge extraction | |
| JPH0512435A (en) | Noise removing method for image | |
| US8180169B2 (en) | System and method for multi-scale sigma filtering using quadrature mirror filters | |
| JP3106831B2 (en) | Video signal processing device | |
| KR101667306B1 (en) | Low-Complexity Topological Derivative-Based Image Segmentation Method and System | |
| JPH1013854A (en) | Video emphasizer device | |
| US5448304A (en) | Process for reducing the noise of a video signal | |
| JPH0927955A (en) | Image processing unit | |
| JPH08317389A (en) | Block distortion remover | |
| JPH04314274A (en) | Video camera with video image blur correction function | |
| JP2001169116A (en) | Method and device for image enhancement | |
| JP2001014474A (en) | Figure extracting method | |
| Vinayagamoorthy et al. | A multichannel filter for TV signal processing | |
| CN110942431B (en) | Image detail enhancement method | |
| JP3463640B2 (en) | Digital noise reduction circuit | |
| KR100222278B1 (en) | Blocking effect removing method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| A300 | Application deemed to be withdrawn because no request for examination was validly filed |
Free format text: JAPANESE INTERMEDIATE CODE: A300 Effective date: 19980903 |