JPH09288736A - Image quality measurement method - Google Patents

Image quality measurement method

Info

Publication number
JPH09288736A
JPH09288736A JP8122816A JP12281696A JPH09288736A JP H09288736 A JPH09288736 A JP H09288736A JP 8122816 A JP8122816 A JP 8122816A JP 12281696 A JP12281696 A JP 12281696A JP H09288736 A JPH09288736 A JP H09288736A
Authority
JP
Japan
Prior art keywords
image
image quality
similarity
processed
images
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.)
Granted
Application number
JP8122816A
Other languages
Japanese (ja)
Other versions
JP3088654B2 (en
Inventor
Hisahiro Tanaka
久博 田中
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nachi Fujikoshi Corp
Original Assignee
Nachi Fujikoshi Corp
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 Nachi Fujikoshi Corp filed Critical Nachi Fujikoshi Corp
Priority to JP08122816A priority Critical patent/JP3088654B2/en
Publication of JPH09288736A publication Critical patent/JPH09288736A/en
Application granted granted Critical
Publication of JP3088654B2 publication Critical patent/JP3088654B2/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Abstract

PROBLEM TO BE SOLVED: To provide an image quality measurement method capable of quantitatively evaluating image quality in the similarity to a source image of an image- processed processing image without being affected by illuminance fluctuation and noise, etc. SOLUTION: In this image quality measurement method for executing an image processing to an image inputted from an ITV camera, that is the source image, and evaluating the image quality in the similarity of the image-processed image, that is the processing image, and the source image, the plural pieces of the images whose image quality is quantitatively evaluated beforehand as the source image are prepared as reference images, normalized coefficients of correlation between the respective plural pieces of the reference images and the image qualities in similarity to respective processed images are calculated by a normalization correlation operation and the mutual centroid operation of the calculated plural pieces of the normalized coefficients of correlation is executed, thus the image quality in the similarity of the source image and the processing image is quantitatively evaluated.

Description

【発明の詳細な説明】Detailed Description of the Invention

【0001】[0001]

【発明の属する技術分野】本発明は、コンピュータなど
を利用して画像のディジタル処理を行うシステムにおい
て、原画像と処理画像との類似度を定量的に測定する画
質測定方法に関する。
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image quality measuring method for quantitatively measuring the degree of similarity between an original image and a processed image in a system for digitally processing an image using a computer or the like.

【0002】[0002]

【従来の技術】原画像f(x,y)に対する処理画像g
(x,y)の類似度、いわゆる画質を定量的に測定する
方法には種々のものが考えだされている。例えば、平均
2乗偏差∬(f−g)2 dxdyを計算する方法、平均
絶対偏差∬|f−g|dxdyを計算する方法、最大絶
対偏差max|f−g|を計算する方法などが提案され
ている。概してこれらの方法は演算方法が比較的簡単な
ため、様々な分野で広く利用されている。また、画像処
理を信号処理の観点から解析し、処理画像g(x,y)
は原画像f(x,y)及び劣化関数h(x,y)の線形
演算から得られるものと仮定し、劣化関数h(x,y)
の特性から処理画像g(x,y)の画質を評価する方法
も提案されており、これについては、例えば長尾真監訳
「ディジタル画像処理」(1978,近代科学社,15
8頁〜171頁)などの文献に詳しく記載されている。
2. Description of the Related Art A processed image g for an original image f (x, y)
Various methods have been devised for quantitatively measuring the degree of similarity of (x, y), so-called image quality. For example, the method of calculating the mean square deviation ∬ (f−g) 2 dxdy, the method of calculating the average absolute deviation ∬ | f−g | dxdy, and the method of calculating the maximum absolute deviation max | f−g | Has been done. Generally, these methods are widely used in various fields due to their relatively simple calculation methods. Further, the image processing is analyzed from the viewpoint of signal processing, and the processed image g (x, y)
Is obtained from the linear operation of the original image f (x, y) and the deterioration function h (x, y), and the deterioration function h (x, y)
A method for evaluating the image quality of the processed image g (x, y) has also been proposed based on the characteristics of, for example, "Digital Image Processing" translated by Shin Nagao (1978, Modern Science Co.
Pages 8 to 171) and the like.

【0003】[0003]

【発明が解決しようとする課題】しかし、平均2乗偏
差、あるいは平均絶対偏差を用いて画質を評価する場合
には、大きな偏差が局所的に存在する場合と小さな偏差
が画像全体に存在する場合とを区別することは困難であ
る。しかも、照度変動のような画像全体にわたる偏差の
影響を排除することは計算原理上不可能である。また、
最大絶対偏差を用いた場合には、雑音等による局所的な
画像の変動が画質の評価に大きく影響する。さらに、劣
化関数h(x,y)を用いて画質を評価する場合には、
解析対象となるのは画像処理システムそのものの性質す
なわちハードウェア自体の特性であり、処理画像g
(x,y)の画質については評価されない。
However, when the image quality is evaluated by using the mean square deviation or the mean absolute deviation, a large deviation locally exists and a small deviation exists in the entire image. It is difficult to distinguish between. Moreover, it is impossible in principle of calculation to eliminate the influence of the deviation over the entire image such as the illuminance variation. Also,
When the maximum absolute deviation is used, the local image fluctuation due to noise or the like greatly affects the image quality evaluation. Furthermore, when the image quality is evaluated using the deterioration function h (x, y),
The target of analysis is the property of the image processing system itself, that is, the property of the hardware itself.
The image quality of (x, y) is not evaluated.

【0004】本発明は上記の問題点を解決するためにな
されたものであり、その目的とするところは、処理画像
g(x,y)の原画像f(x,y)に対する類似度とし
ての画質を、照度変動や雑音等の影響に左右されること
なく定量的に評価することが可能な画質測定方法を提供
することにある。
The present invention has been made in order to solve the above problems, and its purpose is to determine the similarity between the processed image g (x, y) and the original image f (x, y). An object of the present invention is to provide an image quality measuring method capable of quantitatively evaluating the image quality without being affected by the influence of illuminance fluctuation, noise, and the like.

【0005】[0005]

【課題を解決するための手段】上記の目的を達成するた
めに、本発明では、ITVカメラから入力された画像す
なわち原画像に対して画像処理を施し、この画像処理さ
れた画像すなわち処理画像と原画像との類似度としての
画質を評価する画質測定方法において、原画像として予
め画質が定量的に評価されている画像を参照画像として
複数個準備しておき、これら複数個の参照画像のそれぞ
れと処理画像との類似度としての画質を正規化相関係数
として正規化相関演算によりそれぞれ算出し、算出され
た複数個の正規化相関係数相互の重心演算を行うことに
より、原画像と処理画像との類似度としての画質を定量
的に評価するようにした。
In order to achieve the above object, in the present invention, image processing is performed on an image input from an ITV camera, that is, an original image, and the image processed image, that is, a processed image In the image quality measuring method for evaluating the image quality as the degree of similarity with the original image, a plurality of images whose image quality has been quantitatively evaluated in advance are prepared as reference images as reference images, and each of the plurality of reference images is prepared. The image quality as the degree of similarity between the processed image and the processed image is calculated as the normalized correlation coefficient by the normalized correlation operation, and the calculated center of gravity of the normalized correlation coefficients is calculated. The image quality as the similarity to the image is quantitatively evaluated.

【0006】この構成としたことにより、処理画像の原
画像に対する画質の評価は、照度変動の影響を受けるこ
との少ない正規化相関演算により算出された正規化相関
係数を基に導出される。また、原画像として予め画質が
定量的に評価されている画像を参照画像として複数個準
備し、これら複数個の参照画像のそれぞれと処理画像と
の類似度としての画質を算出しているので、雑音が原画
像に混入した場合でも、原画像より得られた複数個の参
照画像のそれぞれにも互いに等しい量の雑音が混入する
ことになり、その結果複数個の参照画像のそれぞれと処
理画像との類似度としての画質の相対的な評価は、雑音
の影響が排除されたものとなる。
With this configuration, the evaluation of the image quality of the processed image with respect to the original image is derived based on the normalized correlation coefficient calculated by the normalized correlation calculation which is less affected by the illuminance change. Further, since a plurality of images whose image quality has been quantitatively evaluated in advance as the original image are prepared as reference images, and the image quality as the similarity between each of the plurality of reference images and the processed image is calculated, Even when noise is mixed in the original image, the same amount of noise is mixed in each of the plurality of reference images obtained from the original image, and as a result, each of the plurality of reference images and the processed image are mixed. The relative evaluation of the image quality as the degree of similarity is such that the influence of noise is eliminated.

【0007】[0007]

【発明の実施の形態】以下、本発明の一実施形態につい
て、図面を参照して説明する。図1は原画像、図2
(a)〜(e)は原画像とITVカメラとの距離を5段
階に変化させながら撮影したときの画像すなわち参照画
像である。表1に示す画質Xk (k=a,b,c,d,
e)は、図2(a)〜(e)のそれぞれの参照画像の、
図1に示す原画像に対する画質(類似度)を主観的に評
価した値である。具体的には、図2(c)の参照画像が
図1に示す原画像に最も類似しているのでこれを基準と
して画質100(Xc =100)と設定し、さらに図2
(c)の参照画像を基準としたときの図2(a)、
(b)、(d)、(e)の各参照画像の画質をそれぞ
れ、30(Xa =30)、50(Xb =50)、130
(Xc =130)、170(Xe =170)と評価し
た。ここでは、画質が100に近いほど原画像に類似し
ている、すなわち画質が良いと評価することになる。
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. Figure 1 is the original image, Figure 2
(A) to (e) are images, that is, reference images when images are taken while changing the distance between the original image and the ITV camera in five steps. Image quality X k (k = a, b, c, d,
e) is the reference image of each of FIGS.
This is a value obtained by subjectively evaluating the image quality (similarity) with respect to the original image shown in FIG. Specifically, since the reference image of FIG. 2C is most similar to the original image shown in FIG. 1, the image quality of 100 (X c = 100) is set with reference to this, and
FIG. 2A when the reference image in FIG.
The image qualities of the reference images of (b), (d), and (e) are set to 30 (X a = 30), 50 (X b = 50), and 130, respectively.
It was evaluated as (X c = 130) and 170 (X e = 170). Here, the closer the image quality is to 100, the more similar to the original image, that is, the better the image quality is evaluated.

【0008】図3は、ITVカメラから入力された画像
すなわち原画像に対して、画像処理装置等により画像処
理を施した後の処理画像の一例を示したものである。こ
の処理画像と図2(a)〜(e)に示す参照画像との正
規化相関係数Ck (k=a,b,c,d,e)を、次式
に示す正規化相関演算により算出する。
FIG. 3 shows an example of a processed image after an image input from an ITV camera, that is, an original image is subjected to image processing by an image processing device or the like. The normalized correlation coefficient C k (k = a, b, c, d, e) between this processed image and the reference image shown in FIGS. 2A to 2E is calculated by the normalized correlation calculation shown in the following equation. calculate.

【0009】[0009]

【数1】 [Equation 1]

【0010】式(1)において、fk (i,j)〔k=
a,b,c,d,e〕は図2に示す各参照画像(a)〜
(e)の各画素の濃度値であり、g(i,j)は図3に
示す原画像の各画素の濃度値である。ここで、(i,
j)は画素の位置を示している。式(1)に示す正規化
相関演算により得られた正規化相関係数Ck は、画像相
互の類似度を評価する指数であり、またこの値は−1〜
+1の間で推移し、+1において最も画像が類似してい
ると評価される。表1は、図2に示した各参照画像と図
3に示した処理画像との正規化相関演算の結果を示した
ものである。
In equation (1), f k (i, j) [k =
a, b, c, d, e] are reference images (a) to FIG.
(E) is the density value of each pixel, and g (i, j) is the density value of each pixel of the original image shown in FIG. Where (i,
j) indicates the position of the pixel. The normalized correlation coefficient C k obtained by the normalized correlation calculation shown in Expression (1) is an index for evaluating the degree of similarity between images, and this value is −1 to −1.
It transitions between +1 and at +1 the images are evaluated as most similar. Table 1 shows the result of the normalized correlation operation between each reference image shown in FIG. 2 and the processed image shown in FIG.

【0011】[0011]

【表1】 [Table 1]

【0012】表1より、図3に示した処理画像との正規
化相関係数が最も高くなったのは図2(c)に示す参照
画像であることがわかる。
It can be seen from Table 1 that the reference image shown in FIG. 2C has the highest normalized correlation coefficient with the processed image shown in FIG.

【0013】次に、次式に示す重心演算を行い、図3に
示した処理画像の画質を算出する。
Next, the center of gravity shown in the following equation is calculated to calculate the image quality of the processed image shown in FIG.

【0014】[0014]

【数2】 [Equation 2]

【0015】表1に示した、各参照画像の原画像に対す
る画質を主観的に評価した値Xk 、及び処理画像と各参
照画像との正規化相関係数Ck を式(2)に代入する
と、図3に示した処理画像の画質Xは91.30と評価
され、図3に示した処理画像は原画像との類似度が比較
的高いと判断される。
The value X k obtained by subjectively evaluating the image quality of each reference image with respect to the original image and the normalized correlation coefficient C k between the processed image and each reference image shown in Table 1 are substituted into the equation (2). Then, the image quality X of the processed image shown in FIG. 3 is evaluated as 91.30, and it is determined that the processed image shown in FIG. 3 has a relatively high similarity to the original image.

【0016】ここで、雑音が原画像に混入した場合を考
えると、原画像より得られた複数個の参照画像のそれぞ
れにも互いに等しい量の雑音が混入することになるの
で、図2(c)の参照画像を基準としたときの図2
(a)、(b)、(d)、(e)の各参照画像の画質の
評価は、雑音の混入がない場合と同一になり、その結果
複数個の参照画像のそれぞれと処理画像との類似度とし
ての画質の評価は、雑音の有無には影響されないものと
なる。
Considering the case where noise is mixed in the original image, the same amount of noise is mixed in each of the plurality of reference images obtained from the original image. 2) when the reference image of FIG.
The evaluation of the image quality of each of the reference images of (a), (b), (d), and (e) is the same as that when no noise is mixed, and as a result, the plurality of reference images and the processed image are compared. The evaluation of the image quality as the similarity is not affected by the presence or absence of noise.

【0017】上記実施形態では、原画像に対する参照画
像の画質Xk は複数の参照画像の任意の一つを基準にし
たときの主観的な評価として表したが、例えば、被撮影
物とITVカメラとの距離をパラメータとして、客間的
な評価として表すようにしてもよい。
In the above embodiment, the image quality X k of the reference image with respect to the original image is expressed as a subjective evaluation when an arbitrary one of the plurality of reference images is used as a reference. It may be expressed as a customer evaluation by using the distance between and as a parameter.

【0018】[0018]

【発明の効果】以上述べたように、本発明では、原画像
として予め画質が定量的に評価されている画像を参照画
像として複数個準備しておき、これら複数個の参照画像
のそれぞれと原画像に対して画像処理を施した処理画像
との類似度としての画質を、正規化相関係数として正規
化相関演算によりそれぞれ算出し、算出された複数個の
正規化相関係数相互の重心演算を行うことにより、原画
像と処理画像との類似度としての画質を定量的に評価す
るようにしたので、処理画像の原画像に対する画質の評
価は、照度変動の影響を受けることの少ない正規化相関
演算により算出された正規化相関係数を基に導出される
こととなった。これにより、照度変動のような画像全体
にわたる偏差の影響を排除できることとなり、例えば、
大きな偏差が局所的に存在する場合と小さな偏差が画像
全体に存在する場合とを区別して評価することも可能と
なった。
As described above, according to the present invention, a plurality of images whose image quality has been quantitatively evaluated in advance are prepared as reference images as reference images, and each of these plurality of reference images and the original image are prepared. The image quality as the degree of similarity to the processed image obtained by performing image processing on the image is calculated by the normalized correlation calculation as the normalized correlation coefficient, and the calculated center of gravity of the plurality of calculated normalized correlation coefficients is calculated. By doing so, the image quality as the degree of similarity between the original image and the processed image is quantitatively evaluated, so the evaluation of the image quality of the processed image with respect to the original image is normalized by the influence of illuminance fluctuation. It will be derived based on the normalized correlation coefficient calculated by the correlation calculation. This makes it possible to eliminate the influence of deviations across the image, such as illuminance fluctuations.
It is also possible to distinguish between cases where large deviations exist locally and cases where small deviations exist throughout the image.

【0019】また、雑音が原画像に混入した場合でも、
原画像より得られた複数個の参照画像のそれぞれにも互
いに等しい量の雑音が混入することになり、その結果複
数個の参照画像のそれぞれと処理画像との類似度として
の画質の評価は、雑音の影響が排除されたものとなった
ので、前処理において雑音を除去する必要はなくなっ
た。
Further, even if noise is mixed in the original image,
The same amount of noise is also mixed in each of the plurality of reference images obtained from the original image, and as a result, the evaluation of the image quality as the similarity between each of the plurality of reference images and the processed image is Since the influence of noise has been eliminated, it is no longer necessary to remove noise in preprocessing.

【図面の簡単な説明】[Brief description of drawings]

【図1】本発明の一実施形態を説明するために用いた原
画像の例を示す図である。
FIG. 1 is a diagram showing an example of an original image used to describe an embodiment of the present invention.

【図2】図1の原画像をITVカメラとの距離を変えて
撮影したときの参照画像の一例を示す図である。
FIG. 2 is a diagram showing an example of a reference image when the original image of FIG. 1 is photographed at different distances from an ITV camera.

【図3】画像処理された処理画像の一例を示す図であ
る。
FIG. 3 is a diagram showing an example of a processed image subjected to image processing.

Claims (1)

【特許請求の範囲】[Claims] 【請求項1】ITVカメラから入力された画像すなわち
原画像に対して画像処理を施し、該画像処理された画像
すなわち処理画像と前記原画像との類似度としての画質
を評価する画質測定方法において、前記原画像として予
め画質が定量的に評価されている画像を参照画像として
複数個準備しておき、該複数個の参照画像のそれぞれと
前記処理画像との類似度としての画質を正規化相関係数
として正規化相関演算により算出し、算出された複数個
の正規化相関係数相互の重心演算を行うことにより、原
画像と処理画像との類似度としての画質を定量的に評価
するようにしたことを特徴とする画質測定方法。
1. An image quality measuring method for performing image processing on an image input from an ITV camera, that is, an original image, and evaluating image quality as a similarity between the image-processed image, that is, the processed image and the original image. A plurality of images whose image quality is quantitatively evaluated beforehand as the original images are prepared as reference images, and the image quality as the similarity between each of the plurality of reference images and the processed image is normalized. Calculate the image quality as the degree of similarity between the original image and the processed image by calculating the correlation number by the normalized correlation calculation and performing the calculation of the center of gravity of the calculated multiple normalized correlation coefficients. An image quality measuring method characterized in that
JP08122816A 1996-04-22 1996-04-22 Image quality measurement method Expired - Fee Related JP3088654B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP08122816A JP3088654B2 (en) 1996-04-22 1996-04-22 Image quality measurement method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP08122816A JP3088654B2 (en) 1996-04-22 1996-04-22 Image quality measurement method

Publications (2)

Publication Number Publication Date
JPH09288736A true JPH09288736A (en) 1997-11-04
JP3088654B2 JP3088654B2 (en) 2000-09-18

Family

ID=14845353

Family Applications (1)

Application Number Title Priority Date Filing Date
JP08122816A Expired - Fee Related JP3088654B2 (en) 1996-04-22 1996-04-22 Image quality measurement method

Country Status (1)

Country Link
JP (1) JP3088654B2 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108109145A (en) * 2018-01-02 2018-06-01 中兴通讯股份有限公司 Picture quality detection method, device, storage medium and electronic device
CN110348314A (en) * 2019-06-14 2019-10-18 中国资源卫星应用中心 A kind of method and system using multi- source Remote Sensing Data data monitoring vegetation growing way
CN113920115A (en) * 2021-12-13 2022-01-11 北京中新绿景科技有限公司 Video image quality evaluation method and system

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108109145A (en) * 2018-01-02 2018-06-01 中兴通讯股份有限公司 Picture quality detection method, device, storage medium and electronic device
CN110348314A (en) * 2019-06-14 2019-10-18 中国资源卫星应用中心 A kind of method and system using multi- source Remote Sensing Data data monitoring vegetation growing way
CN113920115A (en) * 2021-12-13 2022-01-11 北京中新绿景科技有限公司 Video image quality evaluation method and system
CN113920115B (en) * 2021-12-13 2022-03-04 北京中新绿景科技有限公司 Video image quality evaluation method and system

Also Published As

Publication number Publication date
JP3088654B2 (en) 2000-09-18

Similar Documents

Publication Publication Date Title
Shen et al. Hybrid no-reference natural image quality assessment of noisy, blurry, JPEG2000, and JPEG images
US20030112874A1 (en) Apparatus and method for detection of scene changes in motion video
CN108182421A (en) Methods of video segmentation and device
CN111654719B (en) Video micro-motion detection method based on permutation entropy algorithm
CN109344758B (en) Face recognition method based on improved local binary pattern
US20090257673A1 (en) Method and system for noise level detection in image data
CN110186929A (en) A kind of real-time product defect localization method
US20090257675A1 (en) Method and system for noise level detection in image data
US8208753B2 (en) Method and system for noise level detection in image data
JP2875263B2 (en) Image processing method and apparatus
JP3088654B2 (en) Image quality measurement method
JP2002090308A (en) Evaluation system for degree of surface degradation of steel using image processing
CN106611417B (en) Method and device for classifying visual elements into foreground or background
CN115830012A (en) Method for detecting and analyzing contact net clue damage data
DE50310632D1 (en) Objectification of (surface) test methods by image processing
JP2000003436A (en) Device and method for recognizing isar picture
JP2005252451A (en) Image quality inspection method and image quality inspection apparatus
Anitha et al. Quality assessment of resultant images after processing
KR101899070B1 (en) Method and Apparatus for No-reference Image Quality Assessment by Quantfying Robustness against Random Signal
JPH0217832B2 (en)
CN117474915B (en) Abnormality detection method, electronic equipment and storage medium
CN113382134B (en) Focusing debugging method of linear array industrial camera
KR102427597B1 (en) Fine dust detecting solution and system by computing saturation residual based on AI
Castiglione et al. Source camera identification in real practice: A preliminary experimentation
MUSTAFA et al. FLUDI: Using digital images for measuring fish length

Legal Events

Date Code Title Description
A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

Effective date: 20000704

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20070714

Year of fee payment: 7

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20080714

Year of fee payment: 8

LAPS Cancellation because of no payment of annual fees