WO2014107906A1 - 评价标准参数的建立方法及显示屏图像质量的评价方法 - Google Patents
评价标准参数的建立方法及显示屏图像质量的评价方法 Download PDFInfo
- Publication number
- WO2014107906A1 WO2014107906A1 PCT/CN2013/070453 CN2013070453W WO2014107906A1 WO 2014107906 A1 WO2014107906 A1 WO 2014107906A1 CN 2013070453 W CN2013070453 W CN 2013070453W WO 2014107906 A1 WO2014107906 A1 WO 2014107906A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- function
- evaluation
- primary color
- convolution
- parameters
- 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.)
- Ceased
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20056—Discrete and fast Fourier transform, [DFT, FFT]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Definitions
- the present invention relates to the field of image processing, and in particular to a method for establishing evaluation standard parameters, and to an evaluation method for display image quality using evaluation standard parameters.
- the display serves as a friendly interface for human-machine communication and can output information accurately, intuitively and clearly.
- the image quality of the display screen becomes one of the important determinants of consumer purchase.
- the image quality is determined by parameters such as brightness, color shift, and sharpness. Therefore, evaluating the quality of the display screen is an important part of the development and design of the display. Since the display information of the display screen is directly observed by the human eye, the evaluation criteria are different for different environments and uses. In recent years, with the rapid increase in the number and variety of various display technologies, the demand for display image quality has increased.
- the JND standard (Just Noticeable Difference) is used to evaluate the image quality, and it is necessary to determine the JND critical image by using a psychophysical method based visual perception experiment, specifically: each time simultaneously on the display side by side Two images are displayed, one for the original image and one for the test image. Participants were asked to select images from the two images that were considered to have a more pronounced color. Initially, the color deviation between the test image and the original image was very large, and the subjects could easily distinguish the two. If the subject selection is correct, reduce the color deviation between the next test image and the original image; if the subject selects the error, increase the color deviation between the lower test image and the original image. Repeat the above process. When the color deviation does not satisfy certain conditions, the corresponding test image is the JND critical image. For the evaluation of the subsequent images, the JND critical image is used as a reference.
- the results of the image quality evaluation by the naked eye of the subject may differ. Even if the same subject was evaluated, because the actual state of the subjects at different time points was different, the subjects observed that the viewing angle, distance, and lighting conditions of the display screen were different, resulting in the subjects being tested. Subjectivity will be involved in the evaluation, resulting in inconsistent accuracy of the evaluation. In addition, the level of cognition of the severity of color cast in the display will also vary with image brightness and ambient light conditions.
- a primary object of the present invention is to provide a method for establishing an evaluation standard parameter and an evaluation method for the image quality of a display screen, which can replace the observation of the human eye with an evaluation standard parameter.
- another technical solution adopted by the present invention is: providing a method for establishing an evaluation standard parameter, the evaluation standard parameter is used for evaluating the image quality of the display screen, and the image quality is determined by the severity of the image.
- the method of establishing includes: taking a group of test images with different color cast severity to obtain a sample photo group, selecting a standard photo from the sample photo group through the human eye; and performing tristimulus values on the photos in the sample photo group
- the Fourier transform obtains a frequency distribution function corresponding to each primary color, and the frequency distribution functions corresponding to each primary color are respectively expressed as:
- F(i3 ⁇ 4), F(co Y ), 7 ⁇ 3 ⁇ 4 ) are the frequency distribution functions corresponding to the red primary color, the green primary color, and the blue primary color, respectively, F is the action operator of the Fourier transform, and X, Y, and ⁇ represent red,
- ⁇ ( ⁇ ⁇ ) plant F(a> x )CSF(k - ⁇ ) ⁇
- CSF(iy) is the human eye contrast sensitivity function
- ⁇ " ⁇ ⁇ 2 + ⁇ ⁇ 2 is the inter-frequency
- ⁇ 3 ⁇ 4 and ⁇ ⁇ are the other horizontal and vertical spatial frequencies
- ⁇ (), ⁇ ( « ⁇ ) is three types of convolution functions
- k is the offset
- the range is to normalize each type of convolution function separately, and the evaluation parameters of each primary color are obtained.
- the evaluation parameters are expressed as:
- Actor 7 ⁇ actor ⁇ , actor F , and actor z are evaluation parameters; the evaluation parameters of the three primary colors of the standard photograph are selected from the sample photo group as the evaluation standard parameters.
- the steps of normalizing each type of convolution function specifically include: respectively, each type of convolution function is translated as an origin center, and an absolute value of each function of the convolution function at the origin is calculated, and each type of convolution function is calculated.
- the corresponding frequency distribution function corresponds to the sum of the absolute values of the functions of all variables, and the ratio of the absolute value of the function to the sum of the absolute values of the functions.
- the function value of the frequency distribution function corresponding to each primary color at the origin is removed from the sum of the absolute values of the functions.
- the method of establishing includes: taking a group of test images with different color cast severity to obtain a sample photo group, selecting a standard photo from the sample photo group through the human eye; and performing tristimulus values on the photos in the sample photo group
- the Fourier transform the frequency distribution function corresponding to each primary color is obtained; the frequency distribution function corresponding to each primary color is respectively associated with the human eye contrast sensitivity function
- Convolution in the frequency domain three types of convolution functions are obtained; each type of convolution function is normalized to obtain the evaluation parameters of each primary color; the evaluation parameters of the three primary colors of the standard photos are selected from the sample photo group respectively.
- the frequency distribution functions corresponding to each primary color are separately filtered to filter out noise.
- F(i3 ⁇ 4), F(co Y ), 7 ⁇ 3 ⁇ 4 ) are the frequency distribution functions corresponding to the red primary color, the green primary color and the blue primary color, respectively, F is the action operator of the Fourier transform, and X, Y and ⁇ represent The stimuli values of the three primary colors of red, green, and blue, ⁇ is the wavelength, ⁇ ), ⁇ ), ⁇ ) are the spectral irritancy values of the standard chromatic observer, ⁇ ) is the photo-reflectance of the photo, (a, b ) is the visible spectral range, / is the scale factor.
- the human eye contrast sensitivity function is expressed as:
- CSF(iy) is the human eye contrast sensitivity function
- 2 is the spatial frequency
- % is the horizontal and vertical spatial frequency.
- H(w z ) plant F(i3 ⁇ 4 )C ⁇ (k - w)dw
- ⁇ ( ⁇ 3 ⁇ 4), ⁇ ( ⁇ ⁇ ), ⁇ ( ⁇ 3 ⁇ 4 ) are three types of convolution functions, k is an offset, and the range is - ⁇ .
- the steps of normalizing each type of convolution function specifically include: convolving each type The function performs the translation of the origin center respectively, calculates the absolute value of the function of each type of convolution function at the origin, calculates the sum of the absolute values of the functions corresponding to the frequency distribution function corresponding to each type of convolution function, and calculates the absolute value of the function and the absolute value of the function. The ratio of and.
- the function value of the frequency distribution function corresponding to each primary color at the origin is removed from the sum of the absolute values of the functions.
- Actor 7 ⁇ where Factor, actor F , i3 ⁇ 4ctor z are evaluation parameters.
- another technical solution adopted by the present invention is: Providing an evaluation method for image quality of a display screen, and the evaluation method adopts an evaluation standard parameter obtained by any of the above-mentioned establishment methods, and the evaluation method includes: The image is taken to obtain a photo; the Fourier transform of the three stimulus values is obtained, and the frequency distribution function corresponding to each primary color is obtained; the frequency distribution function corresponding to each primary color is respectively associated with the human eye contrast sensitivity function in the frequency domain Convolution is performed to obtain three types of convolution functions; each type of convolution function is separately normalized to obtain evaluation parameters of each primary color; and whether the evaluation parameter of each primary color is greater than the evaluation standard parameter of the corresponding primary color, if there is one The evaluation parameters of the primary colors are larger than the evaluation standard parameters of the corresponding primary colors, and the quality of the image is determined to be unqualified.
- the steps of normalizing each type of convolution function specifically include: respectively, each type of convolution function is translated as an origin center, and an absolute value of each function of the convolution function at the origin is calculated, and each type of convolution function is calculated.
- the corresponding frequency distribution function corresponds to the sum of the absolute values of the functions of all variables, and the ratio of the absolute value of the function to the sum of the absolute values of the functions.
- the method for establishing the evaluation standard parameters of the present invention and the evaluation method for the image quality of the display screen are convoluted by performing a Fourier transform of the three-stimulus value on the image of the image, and then convolving with the human eye contrast-sensitive function.
- the function normalization process obtains the evaluation parameters, and after obtaining all the photos, the standard photo is selected, and the evaluation parameter of the standard photo is the evaluation standard parameter, and the evaluation standard parameter is used instead of the human eye. Observing, it is possible to obtain more objective and systematic evaluation standard parameters, improve the objectivity of image evaluation, and help display system design and development.
- FIG. 1 is a schematic flow chart of a first embodiment of a method for establishing an evaluation standard parameter of the present invention
- FIG. 2 is a view showing an example of a human eye contrast sensitivity function curve
- FIG. 3 is a schematic flow chart of a second embodiment of a method for establishing evaluation standard parameters of the present invention
- FIG. 4 is a schematic flow chart of an embodiment of an evaluation method for image quality of a display screen according to the present invention.
- FIG. 1 is a schematic flow chart of a first embodiment of a method for establishing an evaluation standard parameter according to the present invention.
- Evaluation criteria parameters are used to evaluate the image quality of the display.
- the color shift phenomenon of the image can be directly observed by the human eye, and the image quality is inversely proportional to the severity of the color shift of the image.
- the method of establishing includes the following steps:
- Step S11 Taking a group of test images having different color shift severity to take a picture, obtaining a sample photo group, and selecting a standard photo from the sample photo group through the human eye.
- test images of the group are pure color images, and the severity is arbitrarily distributed.
- the group of tests with arbitrary degree of severity is obtained through artificial design.
- Image The invention does not limit the number of test images of the group, and the number can be selected according to actual test needs.
- test images of the group are photographed by a CCD (Charge-coupled Device). Each test image gets a photo. After the photograph is completed, the sample photo group is obtained.
- CCD Charge-coupled Device
- Step S12 Perform Fourier transform of the tristimulus values on the photos in the sample photo group to obtain a frequency distribution function corresponding to each of the primary colors.
- the solid color picture of the display screen has one color, and matching the color requires a certain number of three primary colors (red, green, blue), and the amount of the three primary colors is the tristimulus value of the color. Therefore, the tristimulus value can indicate the degree of color shift of a picture.
- Fourier transform of the tristimulus value is performed, that is, Fourier transform is performed for each primary color, and three frequency distribution functions can be obtained from each photo. Each frequency distribution function corresponds to one primary color, and the frequency distribution function can be Get some column spectral values.
- Step S13 convolving the frequency distribution function corresponding to each primary color with the human eye contrast sensitivity function in the frequency domain to obtain three types of convolution functions.
- the human eye contrast sensitivity function is a function of the inverse of the detection contrast threshold (called visual acuity) as a function of the spatial frequency of the stimulus.
- Human eye CSF data is measured by human visual psychophysical experiments.
- FIG. 2 is an exemplary diagram of a human eye contrast sensitivity function curve.
- the abscissa indicates the spatial frequency
- the ordinate indicates the visual acuity.
- curves in the figure which are all experimentally conducted by previous scientists. The empirical curves that are close to each other.
- Step S14 normalizing each type of convolution function to obtain evaluation parameters of each primary color.
- the frequency value corresponding to the three types of convolution functions are changed from dimensioned to dimensionless, and the evaluation parameter also indicates the relative value of the spectral value corresponding to the convolution function and the spectral value corresponding to the frequency distribution function. The relationship of values.
- Step S15 selecting the evaluation parameters of the three primary colors of the standard photograph from the sample photo group as the evaluation standard parameters.
- the difference between the evaluation parameters of the standard photos and other evaluation parameters can be directly and clearly obtained from the evaluation parameters, and it can be judged whether the selection of the standard photos is representative and reference.
- the step S11 may be repeated to reselect the standard photograph.
- the evaluation parameter of each stimulus value is used as an evaluation standard parameter, so that the evaluation of the standard parameter is more accurate and systematic.
- the JND critical image is selected, and the JND critical image is used as the standard image.
- the method for establishing the evaluation standard parameter of the present invention calculates the evaluation parameter of the standard photo, selects the evaluation parameter as the evaluation standard parameter, and evaluates the standard parameter as an objective object.
- the evaluation criteria are implemented to evaluate the standard parameters instead of the human eye.
- FIG. 3 is a schematic flow chart of the second embodiment of the method for establishing the evaluation standard parameters of the present invention.
- the method of establishing includes the following steps:
- Step S21 Taking a group of test images with different color shift severity to take a photo, and obtaining a sample photo group.
- Step S22 The standard photo is selected from the sample photo group by the human eye.
- Step S23 Perform Fourier transform of the tristimulus values on the photos in the sample photo group to obtain a frequency distribution function corresponding to each primary color.
- the three stimulus values of the photo are subjected to Fourier transform, and the frequency distribution functions corresponding to each primary color are respectively expressed as:
- F(co x ), F( o Y ), and F( o z ) are the frequency distribution functions corresponding to the red primary color, the green primary color, and the blue primary color, respectively, F is the action operator of the Fourier transform, and X, Y, and ⁇ represent The stimulus values of the three primary colors of red, green, and blue, ⁇ is the wavelength, ⁇ ), ⁇ ), ⁇ ) are the standard chromaticity observer spectral tristimulus values, ⁇ ) is the photo-speech reflectance, (a, b) is the visible spectral range, / is the scale factor.
- Step S24 Filtering the frequency distribution functions corresponding to each primary color separately.
- filtering can filter out noise.
- the frequency distribution function corresponding to each primary color is subjected to mean filtering.
- Step S25 convolving the frequency distribution function corresponding to each primary color with the human eye contrast sensitivity function in the frequency domain to obtain three types of convolution functions.
- the human eye contrast sensitivity function is expressed as:
- CSF(iy) is the human eye contrast sensitivity function
- ⁇ ⁇ ⁇ 2 + is the inter-frequency
- its unit is cycle / degree
- % is the horizontal and vertical spatial frequency.
- the spatial frequency is the number of times the brightness (ie, gray scale) in the unit length changes periodically.
- H(w z ) plant F(i3 ⁇ 4 )C ⁇ (k - w)dw
- ⁇ ( ⁇ 3 ⁇ 4), ⁇ ( ⁇ ⁇ ), ⁇ ( ⁇ 3 ⁇ 4) are three types of convolution functions, k is the offset, and the range is - ⁇ .
- Step S26 Perform each type of convolution function as the origin center translation, and calculate the absolute value of the function of each type of convolution function at the origin.
- the convolution function is used as the origin center translation to translate the center of the convolution function to the coordinate origin, that is, the spectral image corresponding to the convolution function is translated to the coordinate origin, at which the absolute value of the spectral value is the largest, and the maximum frequency
- the absolute value of the value is the absolute value of the function.
- Step S27 Calculate the sum of the absolute values of the functions of the frequency distribution function corresponding to each type of convolution function corresponding to all variables.
- the sum of the absolute values of the functions refers to the sum of the absolute values of the spectral values corresponding to each pixel.
- Step S28 Calculate the ratio of the absolute value of the function to the sum of the absolute values of the functions, and obtain the evaluation parameters of each primary color.
- the sum of the absolute values of the functions includes the absolute value of the spectral value at the origin.
- One photo finally obtains the evaluation parameters of the three primary colors, and the evaluation parameters of each primary color are expressed as:
- Actor 7 ⁇ actor ⁇ , actor F , and actor z are evaluation parameters. This evaluation parameter characterizes the quality of the image.
- the absolute value of the spectral value at the origin is not included in the sum of the absolute values of the function, that is, the sum of the absolute values of the functions excludes the function value of the frequency distribution function at the origin.
- Step S29 The evaluation parameter of the standard photo is selected from the evaluation parameters of all the photos in the sample photo group as the evaluation standard parameter. Among them, after the evaluation parameters of the standard photos are selected, the evaluation parameters of the tristimulus values of X, Y and Z are used as three standard evaluation parameters.
- the method for establishing the evaluation standard parameters of the present embodiment changes the subjective evaluation criteria into an objective evaluation parameter, and the evaluation standard parameters can be used for image evaluation regardless of development, design, test or production.
- FIG. 4 is a schematic flow chart of an embodiment of an evaluation method for image quality of a display screen according to the present invention.
- the evaluation method of the present embodiment adopts the evaluation standard parameters obtained by the establishment method of the foregoing embodiment.
- the evaluation method includes the following steps:
- Step S31 Take a picture of the image of the display screen to obtain a photo.
- Step S32 Perform Fourier transform of the tristimulus value on the photo to obtain a frequency distribution function corresponding to each primary color.
- Step S33 convolving the frequency distribution function corresponding to each primary color with the human eye contrast sensitivity function in the frequency domain to obtain three types of convolution functions.
- Step S34 normalizing each type of convolution function to obtain evaluation parameters of each primary color.
- Step S34 specifically includes: translating the convolution function as the origin center, calculating the absolute value of the function of the convolution function at the origin, calculating the sum of the absolute values of the functions of the frequency distribution function corresponding to all variables, and calculating the absolute value of the function and the absolute value of the function. The ratio of and.
- Step S35 determining whether the evaluation parameter of each primary color is greater than the evaluation standard parameter of the corresponding primary color, and if the evaluation parameter of one primary color is greater than the evaluation standard parameter of the corresponding primary color, determining that the quality of the image is not qualified.
- the evaluation parameters of the red primary color of one photo are compared with the evaluation standard parameters of the red primary color
- the evaluation parameters of the green primary color are compared with the evaluation standard parameters of the green primary color
- the evaluation of the blue primary color is performed.
- the parameter is compared with the evaluation standard parameter of the blue primary color. If the evaluation parameter of any primary color is larger than the evaluation standard parameter of the same primary color, the severity of the color deviation phenomenon is clear. Obviously, the image quality is determined to be unqualified, otherwise it is qualified.
- the method for establishing the evaluation standard parameters of the present invention and the evaluation method for the image quality of the display screen can be evaluated by establishing the evaluation standard parameters, and then comparing the evaluation parameters of any image with the evaluation standard parameters, thereby making the evaluation process more objective. , to reduce the participation of the human eye in the evaluation process, to help display the design and development of the system.
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
Abstract
提供一种评价标准参数的建立方法和显示屏图像质量的评价方法,所述建立方法包括:对具有不同色偏严重程度的一组测试图像进行拍照,获得样本照片组,通过人眼从样本照片组中选定标准照片(S11);对样本照片组中的照片进行三刺激值的傅里叶变换,得到每种原色对应的频率分布函数(S12);将每种原色对应的频率分布函数分别与人眼对比度敏感函数在频域中进行卷积,获得三类卷积函数(S13);对每类卷积函数分别进行归一化处理,得到每种原色的评价参数(S14);以及从样本照片组中选取标准照片的三原色的评价参数分别作为评价标准参数(S15)。
Description
评价标准参数的建立方法及显示屏图像质量的评价方法
【技术领域】
本发明涉及图像处理领域, 特别是涉及一种评价标准参数的建立方法, 还 涉及一种采用评价标准参数的显示屏图像质量的评价方法。
【背景技术】
显示屏作为友好的人机交流信息的界面, 可以准确、 直观、 清晰地输出信 息。 显示屏作为一款电子产品销售时, 显示屏图像质量成为消费者购买的重要 决定因素之一, 图像质量由亮度、 色偏、 清晰度等相关参数决定。 因此, 评价 显示屏图像质量是显示屏开发与设计需要考虑的重要内容。 由于显示屏的显示 信息直接通过人眼观察得到, 对于不同的环境与用途, 其评价标准是不同的。 近年来, 随着各种显示技术的数量和多样性的快速增加, 人们对显示屏图像质 量的要求与日倶增。
现有技术中, 评价图像质量是采用 JND标准 (Just Noticeable Difference, 最小可觉差 ),其需要利用基于心理物理学方法的视觉感知实验确定 JND临界图 像, 具体为: 每次在显示器上同时并排显示两幅图像, 一幅为原始图像, 一幅 为测试图像。 被试需要在两幅图像中选出其认为色偏较明显的图像, 起初, 测 试图像与原始图像间色偏差别很大, 被试可以很容易地将两者区分开。 如果被 试选择正确, 则减小下幅测试图像与原始图像间的色偏差别; 如果被试选择错 误, 则增大下幅测试图像与原始图像的色偏差别。 重复上述过程。 在色偏差别 满足一定条件时, 其所对应的测试图像即为 JND临界图像。 对于之后的图像的 评价, 均以该 JND临界图像为基准。
然而, 由于被试的技术熟练程度不同, 所以这种借助被试肉眼进行的图像 质量评价结果会出现差异。 即使是同一被试进行评价, 由于被试在不同时间点 的实际状态不同, 被试观察显示屏的视角、 距离、 光线条件不同, 导致被试的
主观性会介入到评价中, 导致评价的精确度不一致。 此外, 被试对显示屏中色 偏的严重程度的认知水平也会随图像亮度和周围光线条件的不同而改变。
正如从上述中可以看到的那样, 由于被试的主观性介入到评价中, 所以很 难在显示屏的设计、 开发和消费者之间给出客观和公正的评价标准。
【发明内容】
本发明的主要目的是提供一种评价标准参数的建立方法及显示屏图像质量 的评价方法, 能够以评价标准参数代替人眼的观察。
为解决上述技术问题, 本发明采用的另一个技术方案是: 提供一种评价标 准参数的建立方法, 评价标准参数用于评价显示屏的图像质量, 图像质量的高 低与图像的色偏严重程度成反比, 建立方法包括: 对具有不同色偏严重程度的 一组测试图像进行拍照, 获得样本照片组, 通过人眼从样本照片组中选定标准 照片; 对样本照片组中的照片进行三刺激值的傅里叶变换, 得到每种原色对应 的频率分布函数, 每种原色对应的频率分布函数分别表示为:
(¾ ) = F [Χ] =
J「 Χε-ίωλάλ
J
(¾ ) = F [Z] = [∞ Ζβ-ίωλάλ
J
X = φ{λ)^{λ)άλ
F(i¾)、 F(coY )、 7^¾ )分别为红原色、 绿原色、 蓝原色对应的频率分布函数, F 为傅里叶变换的作用算子, X、 Y、 Ζ代表红、 绿、 蓝三种原色的刺激值, Λ为 波长, μ)、 γμ)、 ^μ)为标准色度观察者光谱三刺激值, ί ^)为照片光语反射 率, (a, b ) 为可见光谱范围, /为比例因子; 对每种原色对应的频率分布函数 分别进行滤波, 以滤除噪声; 将每种原色对应的频率分布函数分别与人眼对比 度敏感函数在频域中进行卷积, 获得三类卷积函数, 人眼对比度敏感函数表示 为:
CSFO) = 2.6 x (0.0192 + 0.114 x x e(- 0 114
三类卷积函数分别表示为:
ϋ(ωχ ) =厂 F(a>x )CSF(k - ω)άω
U(o)y ) = F{ y )CSF{) - ω)άω
Η( wz ) = |∞ Έ{ωζ )CSF (k - ^)d^
CSF(iy)为人眼对比度敏感函数, ω =」ωχ 2 + ωγ 2为 间频率, ί¾禾口 ωγ分另 ll为水平、 垂直方向的空间频率, Η(ί¾)、 Η( )、 Η(«ζ )为三类卷积函数, k为偏移量, 范 围为 对每类卷积函数分别进行归一化处理, 得到每种原色的评价参数, 评价参数表示为:
Factor^ = ~
Ρ(ωχ )
F t (ωζ )
actor7 = ― actor^、 actorF、 actorz为评价参数; 从样本照片组中选取标准照片的三原色的 评价参数分别作为评价标准参数。
其中, 对每类卷积函数分别进行归一化处理的步骤具体包括: 将每类卷积 函数分别做原点中心平移, 计算每类卷积函数在原点的函数绝对值, 计算每类 卷积函数对应的频率分布函数对应所有变量的函数绝对值之和, 计算函数绝对 值与函数绝对值之和的比值。
其中, 计算函数绝对值与函数绝对值之和的比值之前, 从函数绝对值之和 中剔除每种原色对应的频率分布函数在原点的函数值。
为解决上述技术问题, 本发明采用的另一个技术方案是: 提供一种评价标 准参数的建立方法, 评价标准参数用于评价显示屏的图像质量, 图像质量的高 低与图像的色偏严重程度成反比, 建立方法包括: 对具有不同色偏严重程度的 一组测试图像进行拍照, 获得样本照片组, 通过人眼从样本照片组中选定标准 照片; 对样本照片组中的照片进行三刺激值的傅里叶变换, 得到每种原色对应 的频率分布函数; 将每种原色对应的频率分布函数分别与人眼对比度敏感函数
在频域中进行卷积,获得三类卷积函数;对每类卷积函数分别进行归一化处理, 得到每种原色的评价参数; 从样本照片组中选取标准照片的三原色的评价参数 分别作为评价标准参数。
其中, 在将每种原色对应的频率分布函数分别与人眼对比度敏感函数在频 域中进行卷积的步骤之前, 对每种原色对应的频率分布函数分别进行滤波, 以 滤除噪声。
其中, 每种原色对应的频率分布函数分别表示为:
(¾ ) = F [X] = [∞ Xe'^dA
J
J
其中, F(i¾)、 F(coY)、 7^¾ )分别为红原色、 绿原色、 蓝原色对应的频率分 布函数, F为傅里叶变换的作用算子, X、 Y、 Ζ代表红、 绿、 蓝三种原色的刺 激值, Α为波长, μ)、 γμ)、 ^μ)为标准色度观察者光谱三刺激值, ^μ)为照 片光语反射率, (a, b ) 为可见光谱范围, /为比例因子。
其中, 人眼对比度敏感函数表示为:
CSF(«) = 2.6 X (0.0192 + 0.114 X X e(-。 )11
其中, CSF(iy)为人眼对比度敏感函数, 2为空间频率, 和%分 别为水平、 垂直方向的空间频率。
其中, 三类卷积函数分别表示为:
ϋ(ωχ ) = F cox )CSF(k - ω)άω
Η(ωγ ) = F{ y )CSF{) - ω)άω
H(wz ) =厂 F(i¾ )C^(k - w)dw
其中, Η(ί¾)、 Κ(ωγ) , Η(ί¾ )为三类卷积函数, k为偏移量, 范围为 -∞□∞。 其中, 对每类卷积函数分别进行归一化处理的步骤具体包括: 将每类卷积
函数分别做原点中心平移, 计算每类卷积函数在原点的函数绝对值, 计算每类 卷积函数对应的频率分布函数对应所有变量的函数绝对值之和, 计算函数绝对 值与函数绝对值之和的比值。
其中, 计算函数绝对值与函数绝对值之和的比值之前, 从函数绝对值之和 中剔除每种原色对应的频率分布函数在原点的函数值。
其中, 评价参数表示为:
Factor^ = ~
Ρ (ωχ )
Factor,
F t (ωζ )
actor7 = ― 其中, Factor、 actorF , i¾ctorz为评价参数。
为解决上述技术问题, 本发明采用的另一个技术方案是: 提供一种显示屏 图像质量的评价方法, 评价方法采用上述任一种的建立方法得到的评价标准参 数, 评价方法包括: 对显示屏的图像进行拍照, 获得照片; 对照片进行三刺激 值的傅里叶变换, 得到每种原色对应的频率分布函数; 将每种原色对应的频率 分布函数分别与人眼对比度敏感函数在频域中进行卷积, 获得三类卷积函数; 对每类卷积函数分别进行归一化处理, 得到每种原色的评价参数; 判断每种原 色的评价参数是否大于对应原色的评价标准参数, 如果有一种原色的评价参数 大于对应原色的评价标准参数, 确定图像的质量不合格。
其中, 对每类卷积函数分别进行归一化处理的步骤具体包括: 将每类卷积 函数分别做原点中心平移, 计算每类卷积函数在原点的函数绝对值, 计算每类 卷积函数对应的频率分布函数对应所有变量的函数绝对值之和, 计算函数绝对 值与函数绝对值之和的比值。
综上所述, 本发明的评价标准参数的建立方法及显示屏图像质量的评价方 法通过对图像的照片进行三刺激值的傅里叶变换, 然后与人眼对比度敏感函数 卷积,将卷积函数归一化处理得到评价参数,在获得所有照片后选定标准照片, 该标准照片的评价参数即为评价标准参数, 实现了以评价标准参数代替人眼的
观察, 能够得到较为客观和系统的评价标准参数, 提高图像评价的客观性, 帮 助显示系统的设计与开发。
上述说明仅是本发明技术方案的概述, 为了能够更清楚了解本发明的技术 手段,而可依照说明书的内容予以实施,并且为了让本发明的上述和其他目的、 特征和优点能够更明显易懂, 以下特举较佳实施例, 并配合附图, 详细说明如 下。
【附图说明】
图 1是本发明评价标准参数的建立方法第一实施例的流程示意图; 图 2是一种人眼对比度敏感函数曲线的示例图;
图 3是本发明评价标准参数的建立方法第二实施例的流程示意图; 图 4是本发明显示屏图像质量的评价方法实施例的流程示意图。
【具体实施方式】
下面将结合本发明实施例中的附图, 对本发明实施例中的技术方案进行清 楚、 完整地描述, 显然, 所描述的实施例仅仅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做出创造 性劳动前提下所获得的所有其他实施例, 均属于本发明保护的范围。
请参阅图 1 ,图 1是本发明评价标准参数的建立方法第一实施例的流程示意 图。 评价标准参数用于评价显示屏的图像质量。 本发明实施例中, 图像的色偏 现象能够被人眼直接观察得到, 图像质量的高低与图像的色偏严重程度成反比。 建立方法包括以下步骤:
步骤 S11: 对具有不同色偏严重程度的一组测试图像进行拍照, 获得样本照 片组, 通过人眼从样本照片组中选定标准照片。
其中, 该组测试图像为纯色图像, 且具有的严重程度是任意分布的, 在开 发与设计显示屏的过程中, 通过人为设计而得到严重程度任意分布的该组测试
图像。 本发明并不对该组测试图像的数量做限制, 数量可以根据实际测试需要 选取。
该组测试图像通过 CCD( Charge-coupled Device,电荷耦合元件 M义器拍照。 每幅测试图像得到一张照片, 拍照完成后, 获得样本照片组。
现有的显示屏和压缩编码等相关技术还不能做到使图像质量非常完美, 总 是存在不尽如人意的缺陷。 由于影响最终图像质量的各技术参数间存在交互影 响, 提高各参数的成本也存在差异, 而成本是实际生产中必须考虑的问题, 因 而改善图像质量的开发与设计需要考虑各因素的权重与折中。 考虑到显示屏的 图像质量是由消费者通过人眼直接观察进行判断的, 而人眼对图像具有的严重 程度的感知能力有限,有的图像存在一定的严重程度,但人眼却并不能感知到, 可以认为这样的图像质量合格。 因此, 在本实施例中, 通过人眼选定一幅照片 作为标准照片, 该照片的严重程度几乎不影响图像的质量, 能够被人眼接受。
步骤 S12: 对样本照片组中的照片进行三刺激值的傅里叶变换,得到每种原 色对应的频率分布函数。
其中, 显示屏的纯色画面具有一种颜色, 而匹配该颜色需要一定数量的三 原色(红、 绿、 蓝), 三原色的量为该颜色的三刺激值。 因此, 三刺激值可以表 征一幅图片的色偏程度。 进行三刺激值的傅里叶变换, 即是对每个原色进行傅 里叶变换, 从每张照片能够得到三个频率分布函数, 每个频率分布函数对应一 种原色, 从频率分布函数上可以得到一些列频谱值。
步骤 S13:将每种原色对应的频率分布函数分别与人眼对比度敏感函数在频 域中进行卷积, 获得三类卷积函数。
其中, 人眼对比度敏感函数( CSF, contrast sensitivity function )为探测对比 度阈值的倒数(称作视觉敏锐度) 随刺激的空间频率变化的函数。 人眼 CSF数 据是通过人眼视觉心理物理学实验测量得到。 例如, 请参阅图 2, 图 2是一种人 眼对比度敏感函数曲线的示例图。 图中, 横坐标表示空间频率, 纵坐标表示视 觉敏锐度。 图中具有若干条曲线, 其均是由之前的科技工作者经过一些实验得
到的相互靠近的经验曲线。
每一种原色对应的频率分布函数与人眼对比度敏感函数进行卷积后, 就得 到一类卷积函数, 该频率分布函数上原来的每个频谱值相应改变, 即重新得到 一组频语值, 这些频语值均可以从该类卷积函数上得到。
步骤 S14:对每类卷积函数分别进行归一化处理,得到每种原色的评价参数。 其中,归一化处理后,三类卷积函数对应的频语值均由有量纲变为无量纲, 评价参数也就表示卷积函数对应的频谱值与频率分布函数对应的频谱值的相对 值的关系。
步骤 S15:从样本照片组中选取标准照片的三原色的评价参数分别作为评价 标准参数。
其中, 所有的评价参数得到后, 从评价参数中可以直接明了地获知标准照 片的评价参数与其它评价参数的差别, 可以判断标准照片的选取是否具有代表 性和参照性。
如果认为标准照片的评价参数的参照性的选取存在较大误差, 可以重复步 骤 S11 , 重新选择出标准照片。 这里, 由于图像的色偏由三刺激值确定, 因此, 每个刺激值的评价参数都作为一种评价标准参数, 这样评价标准参数才会更准 确和系统。
相较现有技术选择 JND临界图像, 以 JND临界图像作为标准图像而言, 本 发明的评价标准参数的建立方法计算标准照片的评价参数, 选取该评价参数作 为评价标准参数, 评价标准参数作为客观评价标准, 实现以评价标准参数代替 人眼的观察。
请参阅图 3,图 3是本发明评价标准参数的建立方法第二实施例的流程示意 图。 建立方法包括以下步骤:
步骤 S21: 对具有不同色偏严重程度的一组测试图像进行拍照, 获得样本照 片组。
步骤 S22: 通过人眼从样本照片组中选定标准照片。
其中, 步骤 S21和步骤 S22请参照前述实施例, 此处不再赘述。 步骤 S23: 对样本照片组中的照片进行三刺激值的傅里叶变换,得到每种原 色对应的频率分布函数。
其中, 这里对照片的三个刺激值都进行傅里叶变换, 每种原色对应的频率 分布函数分别表示为:
F(cox )、 F( oY )、 F( oz )分别为红原色、绿原色、蓝原色对应的频率分布函数, F为傅里叶变换的作用算子, X、 Y、 Ζ代表红、 绿、 蓝三种原色的刺激值, Λ为 波长, μ)、 γμ)、 ^μ)为标准色度观察者光谱三刺激值, ί^μ)为照片光语反射 率, (a, b ) 为可见光谱范围, /为比例因子。
步骤 S24: 对每种原色对应的频率分布函数分别进行滤波。
其中, 滤波可以滤除噪声。 在本实施例中, 对每种原色对应的频率分布函 数进行均值滤波。
步骤 S25:将每种原色对应的频率分布函数分别与人眼对比度敏感函数在频 域中进行卷积, 获得三类卷积函数。
其中, 由于三刺激值都对应有频率分布函数, 因此卷积需要进行三次。 人 眼对比度敏感函数表示为:
CSF(ω) = 2.6 X (0.0192 + 0.114 X X e(-。 )11
CSF(iy)为人眼对比度敏感函数, ω = ^ωχ 2 + 为 间频率 , 其单位为周期 / 度, 和%分别为水平、垂直方向的空间频率。空间频率是指单位长度内亮度(也 就是灰度)作周期性变化的次数。
三类卷积函数分别表示为:
ϋ(ωχ ) = F cox )CSF(k - ω)άω
Η(ωγ ) = F{ Y)CSF{) - ω)άω
H(wz ) =厂 F(i¾ )C^(k - w)dw
Η(ί¾)、 Κ(ωγ) , Η(ί¾)为三类卷积函数, k为偏移量, 范围为 -∞□∞。
步骤 S26: 将每类卷积函数分别做原点中心平移,计算每类卷积函数在原点 的函数绝对值。
其中, 将卷积函数做原点中心平移是指将卷积函数的中心平移至坐标原点, 即将卷积函数对应的频谱图像平移至坐标原点,在该原点,频谱值绝对值最大, 而该最大频语值的绝对值就是函数绝对值。
步骤 S27:计算每类卷积函数对应的频率分布函数对应所有变量的函数绝对 值之和。
其中, 函数绝对值之和是指每个像素对应的频谱值的绝对值之和。
步骤 S28: 计算函数绝对值与函数绝对值之和的比值,得到每种原色的评价 参数。
其中, 在本实施例中, 在计算时, 函数绝对值之和中包括原点处的频谱值 的绝对值。一幅照片最后得到三原色的评价参数,每种原色的评价参数表示为:
Factor^ = ~
Ρ(ωχ )
F t (ωζ )
actor7 = ― actor^、 actorF、 actorz为评价参数。 该评价参数表征了图像质量的高低。 在其它实施例中, 函数绝对值之和中不包括原点处的频谱值的绝对值, 即函数 绝对值之和中剔除了频率分布函数在原点的函数值。
步骤 S29:从样本照片组中所有照片的评价参数中选取标准照片的评价参数 作为评价标准参数。
其中, 选定标准照片的评价参数后, 其 X、 Y和 Z的三刺激值的评价参数 作为三种标准评价参数。
本实施例的评价标准参数的建立方法将主观性的评价标准变为客观化的评 价参数, 无论对于开发、 设计、 测试或生产, 利用该评价标准参数对于图像评 价都能提供帮助。
请参阅图 4 ,图 4是本发明显示屏图像质量的评价方法实施例的流程示意图。 本实施例的评价方法采用前述实施例的建立方法得到的评价标准参数。 评价方 法包括以下步骤:
步骤 S31 : 对显示屏的图像进行拍照, 获得照片。
步骤 S32: 对照片进行三刺激值的傅里叶变换,得到每种原色对应的频率分 布函数。
步骤 S33:将每种原色对应的频率分布函数分别与人眼对比度敏感函数在频 域中进行卷积, 获得三类卷积函数。
其中, 步骤 S31、 S32以及 S33请参照前述实施例的相关步骤, 此处不再赘 述。
步骤 S34:对每类卷积函数分别进行归一化处理,得到每种原色的评价参数。 其中, 步骤 S34具体包括: 将卷积函数做原点中心平移, 计算卷积函数在 原点的函数绝对值, 计算频率分布函数对应所有变量的函数绝对值之和, 计算 函数绝对值与函数绝对值之和的比值。
步骤 S35: 判断每种原色的评价参数是否大于对应原色的评价标准参数, 如 果有一种原色的评价参数大于对应原色的评价标准参数, 确定图像的质量不合 格。
其中, 以标准照片的评价标准参数为基准, 将一幅照片红原色的评价参数 与红原色的评价标准参数作比较、 绿原色的评价参数与绿原色的评价标准参数 作比较以及蓝原色的评价参数与蓝原色的评价标准参数作比较, 如果有任一原 色的评价参数大于其相同原色的评价标准参数, 则说明色偏现象的严重程度明
显, 确定图像质量不合格, 否则合格。
通过上述方式, 本发明的评价标准参数的建立方法及显示屏图像质量的评 价方法通过建立评价标准参数, 再将任意图像的评价参数与评价标准参数作比 较进行评价, 能够使评价过程更为客观, 减少评价过程中人眼的参与, 有助于 显示系统的设计与开发。
以上所述仅为本发明的实施例, 并非因此限制本发明的专利范围, 凡是利 用本发明说明书及附图内容所作的等效结构或等效流程变换, 或直接或间接运 用在其他相关的技术领域, 均同理包括在本发明的专利保护范围内。
Claims
1、 一种评价标准参数的建立方法, 所述评价标准参数用于评价显示屏的图 像质量, 图像质量的高低与图像的色偏严重程度成反比, 其中, 所述建立方法 包括:
对具有不同色偏严重程度的一组测试图像进行拍照, 获得样本照片组, 通 过人眼从所述样本照片组中选定标准照片;
对所述样本照片组中的照片进行三刺激值的傅里叶变换, 得到每种原色对 应的频率分布函数, 所述每种原色对应的频率分布函数分别表示为:
(¾) = F [X]= [∞ Xe'^dA
J
J
F(a?x)、 F(a)Y)、 7^¾)分别为红原色、 绿原色、 蓝原色对应的频率分布函数,
F为傅里叶变换的作用算子, X、 Y、 Ζ代表红、 绿、 蓝三种原色的刺激值, Λ为 波长, μ)、 γμ)、 ^μ)为标准色度观察者光谱三刺激值, ί^μ)为照片光语反射 率, (a, b) 为可见光谱范围, /为比例因子;
对每种原色对应的所述频率分布函数分别进行滤波, 以滤除噪声; 将每种原色对应的所述频率分布函数分别与人眼对比度敏感函数在频域中 进行卷积, 获得三类卷积函数, 所述人眼对比度敏感函数表示为:
CSF(«) = 2.6 X (0.0192 + 0.114 X X e(-。 )11
所述三类卷积函数分别表示为:
Η(ωχ) = F cox )CSF(k - ω)άω
Η(ωγ) = F{ y)CSF{) - ω)άω
H(wz ) =厂 F(i¾ )C^(k - w)dw
CSF(iy)为所述人眼对比度敏感函数, 《 = ^^ 7^ 为空间频率, 和%分别 为水平、 垂直方向的空间频率, Η(ί¾)、 Η( )、 ¾ )为所述三类卷积函数, k 为偏移量, 范围为—00□ 00;
对每类所述卷积函数分别进行归一化处理, 得到每种原色的评价参数, 所 述评价参数表示为:
r Η(ωχ )
Factor^ = ~
F t (ωζ )
actor7 = ― actor^ , i¾ctorr、 i¾ctorz为所述评价参数;
从所述样本照片组中选取所述标准照片的三原色的评价参数分别作为评价 标准参数。
2、 根据权利要求 1所述的建立方法, 其中, 所述对每类所述卷积函数分别 进行归一化处理的步骤具体包括:
将每类所述卷积函数分别做原点中心平移, 计算每类所述卷积函数在原点 的函数绝对值, 计算每类所述卷积函数对应的所述频率分布函数对应所有变量 的函数绝对值之和, 计算所述函数绝对值与所述函数绝对值之和的比值。
3、 根据权利要求 2所述的建立方法, 其中, 计算所述函数绝对值与所述函 数绝对值之和的比值之前, 从所述函数绝对值之和中剔除每种原色对应的所述 频率分布函数在所述原点的函数值。
4、 一种评价标准参数的建立方法, 所述评价标准参数用于评价显示屏的图 像质量, 图像质量的高低与图像的色偏严重程度成反比, 其中, 所述建立方法 包括:
对具有不同色偏严重程度的一组测试图像进行拍照, 获得样本照片组, 通 过人眼从所述样本照片组中选定标准照片;
对所述样本照片组中的照片进行三刺激值的傅里叶变换, 得到每种原色对 应的频率分布函数;
将每种原色对应的所述频率分布函数分别与人眼对比度敏感函数在频域中 进行卷积, 获得三类卷积函数;
对每类所述卷积函数分别进行归一化处理, 得到每种原色的评价参数; 从所述样本照片组中选取所述标准照片的三原色的评价参数分别作为评价 标准参数。
5、 根据权利要求 4所述的建立方法, 其中, 在将每种原色对应的所述频率 分布函数分别与人眼对比度敏感函数在频域中进行卷积的步骤之前, 对每种原 色对应的所述频率分布函数分别进行滤波, 以滤除噪声。
6、 根据权利要求 4所述的建立方法, 其中, 所述每种原色对应的频率分布 函数分别表示为:
(¾ ) = F [Χ] =「 Χε-ίωλάλ
J
J
(¾ ) = F [Z] = [∞ Ζβ-ίωλάλ
J
Y = φ{λ)^{λ)άλ
其中, F(i¾)、 F(a?r)、 7^¾ )分别为红原色、 绿原色、 蓝原色对应的频率分 布函数, F为傅里叶变换的作用算子, X、 Y、 Ζ代表红、 绿、 蓝三种原色的刺 激值, Α为波长, (λ) , γμ)、 ^μ)为标准色度观察者光谱三刺激值, ^μ)为照 片光语反射率, (a, b ) 为可见光谱范围, /为比例因子。
7、 根据权利要求 6所述的建立方法, 其中, 所述人眼对比度敏感函数表示 为:
CSF(«) = 2.6 X (0.0192 + 0.114 X X e(- 0 114 1 1
其中 , CSF(iy)为所述人目艮对比度敏感函数, ω =」ωχ 2 + ωγ 2为空间频率, %和% 分别为水平、 垂直方向的空间频率。
8、根据权利要求 7所述的建立方法,其中,所述三类卷积函数分别表示为:
ϋ(ωχ ) = F{ x )CSF(k - ω)άω
Η(ω, ) = |∞ F{ y )CSF{) - ω)άω
Η(ί¾ ) = |∞ Έ{ωζ )CSF(k - ω)άω 其中, Η(ί¾)、 Η(%)、 ¾(¾)为所述三类卷积函数, k为偏移量,范围为 -∞□∞。
9、 根据权利要求 8所述的建立方法, 其中, 所述对每类所述卷积函数分别 进行归一化处理的步骤具体包括: 将每类所述卷积函数分别做原点中心平移, 计算每类所述卷积函数在原点 的函数绝对值, 计算每类所述卷积函数对应的所述频率分布函数对应所有变量 的函数绝对值之和, 计算所述函数绝对值与所述函数绝对值之和的比值。
10、 根据权利要求 9所述的建立方法, 其中, 计算所述函数绝对值与所述 函数绝对值之和的比值之前, 从所述函数绝对值之和中剔除每种原色对应的所 述频率分布函数在所述原点的函数值。
11、 根据权利要求 9所述的建立方法, 其中, 所述评价参数表示为:
Factor^ = ~
Ρ(ωχ )
F t (ωζ )
actor7 = ― 其中, Factor、 actorF , i¾ctorz为所述评价参数。
12、 一种显示屏图像质量的评价方法, 其中, 所述评价方法采用根据权利 要求 4所述的建立方法得到的评价标准参数, 所述评价方法包括:
对所述显示屏的图像进行拍照, 获得照片;
对所述照片进行三刺激值的傅里叶变换, 得到每种原色对应的频率分布函 数;
将每种原色对应的所述频率分布函数分别与人眼对比度敏感函数在频域中 进行卷积, 获得三类卷积函数;
对每类所述卷积函数分别进行归一化处理, 得到每种原色的评价参数; 判断每种原色的评价参数是否大于对应原色的所述评价标准参数, 如果有
一种原色的评价参数大于对应原色的所述评价标准参数, 确定所述图像的质量 不合格。
13、 根据权利要求 12所述的评价方法, 其特征在于, 所述对每类所述卷积 函数分别进行归一化处理的步骤具体包括:
将每类所述卷积函数分别做原点中心平移, 计算每类所述卷积函数在原点 的函数绝对值, 计算每类所述卷积函数对应的所述频率分布函数对应所有变量 的函数绝对值之和, 计算所述函数绝对值与所述函数绝对值之和的比值。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US13/812,879 US8958639B2 (en) | 2013-01-12 | 2013-01-15 | Method for establishing evaluation standard parameters and method for evaluating the quality of a display image |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310011245.X | 2013-01-12 | ||
| CN201310011245.XA CN103049915B (zh) | 2013-01-12 | 2013-01-12 | 评价标准参数的建立方法及显示屏图像质量的评价方法 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2014107906A1 true WO2014107906A1 (zh) | 2014-07-17 |
Family
ID=48062544
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2013/070453 Ceased WO2014107906A1 (zh) | 2013-01-12 | 2013-01-15 | 评价标准参数的建立方法及显示屏图像质量的评价方法 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN103049915B (zh) |
| WO (1) | WO2014107906A1 (zh) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103500457B (zh) * | 2013-10-23 | 2016-06-22 | 武汉东智科技有限公司 | 一种视频图像偏色检测的方法 |
| CN105915791B (zh) * | 2016-05-03 | 2019-02-05 | Oppo广东移动通信有限公司 | 电子装置控制方法及装置、电子装置 |
| CN106409193A (zh) * | 2016-07-08 | 2017-02-15 | 深圳市华星光电技术有限公司 | 一种显示屏残影检测系统及其方法 |
| CN107371015A (zh) * | 2017-07-21 | 2017-11-21 | 华侨大学 | 一种无参考对比度变化图像质量评价方法 |
| CN107590802B (zh) * | 2017-09-11 | 2021-09-14 | 康佳集团股份有限公司 | 一种电视机显示一致性检测方法、存储介质及检测设备 |
| CN110378863A (zh) * | 2018-04-10 | 2019-10-25 | 深圳Tcl新技术有限公司 | 一种显示屏的均匀性检测方法、系统及存储介质 |
| CN108932936B (zh) * | 2018-09-10 | 2020-12-04 | 深圳市华星光电技术有限公司 | 灰阶亮度的调节方法及显示装置 |
| CN113539205B (zh) * | 2021-07-30 | 2022-07-29 | 深圳市华星光电半导体显示技术有限公司 | 像素电压的确定方法、装置、电子设备以及存储介质 |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1156371C (zh) * | 2001-11-20 | 2004-07-07 | 鹤山雅图仕印刷有限公司 | 光栅片与影像线条合成的套色检测方法 |
| CN100465989C (zh) * | 2007-10-09 | 2009-03-04 | 天津工业大学 | 基于综合特征一致性模型的人造目标检测方法 |
| CN102855631A (zh) * | 2012-08-23 | 2013-01-02 | 浙江大学 | 用于图像质量评价的视觉能量信息提取方法 |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2077527A1 (en) * | 2008-01-04 | 2009-07-08 | Thomson Licensing | Method for assessing image quality |
| CN102142145B (zh) * | 2011-03-22 | 2012-12-19 | 宁波大学 | 一种基于人眼视觉特性的图像质量客观评价方法 |
-
2013
- 2013-01-12 CN CN201310011245.XA patent/CN103049915B/zh not_active Expired - Fee Related
- 2013-01-15 WO PCT/CN2013/070453 patent/WO2014107906A1/zh not_active Ceased
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1156371C (zh) * | 2001-11-20 | 2004-07-07 | 鹤山雅图仕印刷有限公司 | 光栅片与影像线条合成的套色检测方法 |
| CN100465989C (zh) * | 2007-10-09 | 2009-03-04 | 天津工业大学 | 基于综合特征一致性模型的人造目标检测方法 |
| CN102855631A (zh) * | 2012-08-23 | 2013-01-02 | 浙江大学 | 用于图像质量评价的视觉能量信息提取方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103049915B (zh) | 2016-01-06 |
| CN103049915A (zh) | 2013-04-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2014107906A1 (zh) | 评价标准参数的建立方法及显示屏图像质量的评价方法 | |
| WO2014107909A1 (zh) | 评价标准参数的建立方法及显示屏图像质量的评价方法 | |
| TWI627854B (zh) | Uneven spot inspection system, uneven spot inspection method and uneven spot inspection program | |
| JP5635677B2 (ja) | 高ダイナミックレンジ、視覚ダイナミックレンジおよび広色範囲の画像およびビデオの品質評価 | |
| Hubbard et al. | Brightness, contrast, and color balance of digital versus film retinal images in the age-related eye disease study 2 | |
| CN102629379A (zh) | 一种基于视觉特性的图像质量评价方法 | |
| WO2011121760A9 (ja) | 画像処理装置、およびそれを用いた撮像装置 | |
| CN106023111A (zh) | 一种影像融合质量评价方法及系统 | |
| US8958639B2 (en) | Method for establishing evaluation standard parameters and method for evaluating the quality of a display image | |
| Hanhart et al. | HDR image compression: a new challenge for objective quality metrics | |
| US8989488B2 (en) | Method for establishing an evaluation standard parameter and method for evaluating the quality of a display image | |
| JP2017051474A (ja) | 肌の光沢評価装置、光沢評価方法および光沢評価プログラム | |
| Pinson | Why no reference metrics for image and video quality lack accuracy and reproducibility | |
| CN116862904B (zh) | 一种基于最小可觉差的显示面板Mura缺陷全局评估方法 | |
| US20170214911A1 (en) | Image-sticking inspection method for liquid crystal panel and device of the same | |
| CN104270579A (zh) | 一种终端应用界面像素调整方法及终端 | |
| TWI755390B (zh) | 不均勻性評估方法及不均勻性評估裝置 | |
| Jin et al. | Softcopy quality ruler method: implementation and validation | |
| JP5163940B2 (ja) | 画質検査装置および画質検査方法 | |
| TWI267737B (en) | Method and device for detecting flat panel display device by visual model | |
| JP5321939B2 (ja) | 画質検査装置および画質検査方法 | |
| Liu et al. | Quality assessment for out-of-focus blurred images | |
| KR101915039B1 (ko) | 색각이상자를 위한 자동 색상 보정 방법 및 장치 | |
| Ponomarenko et al. | Statistical evaluation of no-reference image visual quality metrics | |
| Xu et al. | A database of visual color differences of modern smartphone photography |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| WWE | Wipo information: entry into national phase |
Ref document number: 13812879 Country of ref document: US |
|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13870662 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: 13870662 Country of ref document: EP Kind code of ref document: A1 |