JPH095245A - Method for determining image quality - Google Patents

Abstract

PURPOSE: To judge an inspection result from a total view point, by inspecting the quality of an inspection image correspondingly to inspection items by an inspector and converting the inspection result into a score. CONSTITUTION: An object 5 to be inspected is placed on a transparent document stage 2 in a darkroom 4. When the object 5 is permeable, a passing light, lamp 1 emits light to illuminate the object 5 through the stage 2. The passing light forms an image on a photodetecting element of a color camera 6. The camera 6 converts the image to R, G, B electric signals by red, green, blue (R, G, B) filters. The data is converted to a data fit, to an inspecting means, e.g., Y, x, y data of Yxy coordinates, and stored in a computer 7. The object is inspected correspondingly to inspection items by an inspector with the use of the stored data. When a luminance is inspected, a predetermined standard value and the whole Y data of luminance are compared, a score of a rank including the Y data of the inspection image is set as a score of the Y data, and an average score is set as a score of the luminance inspection. Inspection results of a chromaticity, a color difference, an irregularity are similarly expressed by scores. The image quality is totally judged from the total of scores.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of judging image quality in an image quality inspection.

[0002]

2. Description of the Related Art Conventionally, image quality is judged based on a visual inspection by an inspector or a partial measurement by a color difference meter.
The inspector sensually made a comprehensive decision.

[0003]

However, the conventional method by the inspector has a problem that the inspector's subjectivity is involved in the determination, and the determination varies from one inspector to another, so that a quantitatively stable determination cannot be made. . Then, this invention aims at the following. (1) Comprehensive and comprehensive support for inspectors to inspect images by a plurality of inspection means corresponding to a plurality of viewpoints An object is to provide an image quality judgment method having a quantitative judgment means. (2) To provide an image quality judgment method having means for setting a weight value for each inspection means in order to correspond to the inspector's subjectivity and making a composite and quantitative determination with the inspection result.

[0004]

According to the present invention, in order to solve the conventional technique in which an inspector cannot make a quantitative and stable determination, one image to be inspected is inspected by a plurality of inspecting means, and the inspecting means is inspected. In the image quality judgment method for judging the image quality by the comprehensive judgment means based on a plurality of inspection results obtained from the above, in the plural judgment means, each of the inspection results is expressed as a score in an arbitrarily specified range, and the comprehensive judgment means is used. The image quality is determined by calculating the score by an arbitrary calculation method and obtaining a total score.

Further, the image quality is judged by setting weight values for a plurality of inspection means within an arbitrarily standardized range, and calculating the score and weight value by an arbitrary calculation method by the comprehensive judgment means to obtain a total score. It is characterized by performing. Furthermore, the result obtained by inspecting one or more inspecting means with respect to a plurality of images displayed on the inspecting target display device is expressed as a score in an arbitrarily standardized range, and the score is arbitrarily determined by the comprehensive determining means. The image quality of the display device to be inspected is determined by performing the calculation according to the above calculation method to obtain a total score.

Further, the weight value for a plurality of inspection images and the weight value for a plurality of inspection means are set within a standard range, and the score and weight value obtained by the inspection means are calculated by an arbitrary calculation method to obtain a total score. It is characterized in that the image quality is judged by obtaining it.

[0007]

According to the image quality judging method, the image quality is judged by the comprehensive judging means on the basis of a plurality of inspection results obtained from the inspecting means for inspecting one image to be inspected. Since each test result represents a score in an arbitrarily specified range, it is possible to handle a plurality of test results with the same standard and with a quantitative value. In addition, since the total determination unit calculates the score by an arbitrary calculation method to obtain the total score, it is possible to express the image quality of the target image by a quantitative value.

Further, the weight value is inspected in order to set a weight value for a plurality of inspection means within an arbitrarily standardized range, and to calculate the score and the weight value by an arbitrary calculation method by the comprehensive judgment means to obtain a total score. By setting a value corresponding to the subjectivity of the worker, the image quality determination can be performed according to the subjectivity of the inspector.

Further, it is a method for judging the image quality of a display device, in which a plurality of images corresponding to an inspection item are inspected by one or a plurality of inspecting means, and the obtained results are scored within an arbitrarily specified range. Representation, by comprehensive judgment means,
Since the score is calculated by an arbitrary calculation method to obtain the total score, it is possible to comprehensively and quantitatively judge a large number of test results obtained by the tests performed in combination.

Further, weight values for a plurality of inspection images and weight values for a plurality of inspection means are set within a standard range, and the scores and weight values obtained by the inspection means are calculated by an arbitrary calculation method to obtain a total score. Therefore, by setting a value corresponding to the subjectivity of the inspector for the image quality of the display device, it is possible to determine the image quality that corresponds to the subjectivity of the inspector and is quantitative and stable.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described below with reference to the drawings. FIG. 2 shows an embodiment of an image quality inspection system to which the image quality determination method according to the present invention is applied.

In FIG. 2, an inspection object 5 is installed on the original table 2 inside the dark room 4. The document table 2 is made of a transparent object such as glass, and when the inspection object 5 is a transmissive object, the transmission lamp 1 emits light, and the inspection object 5 is irradiated through the document table 2. The transmitted light of the inspection object 5 is imaged on the light receiving element of the color camera 6 by the lens of the color camera 6 installed above the dark room 4. When the inspection object 5 is a reflective object, the reflection lamp 3 emits light and the inspection object 5
The reflected light reflected by is imaged on the color camera 6.
The color camera 6 is equipped with red, green, and blue (hereinafter abbreviated as R, G, B) filters, and converts an image of the inspection object 5 into R, G, B electrical signals and outputs the electrical signals. Further, the electric signal is converted into an R, G, B digital signal by an A / D converter (not shown) and transferred to the computer 7.

In this embodiment, the color camera 6 is used to obtain a color image, but a monochrome camera may be combined with a color filter or a color light source. Further, when the determination process is performed using the brightness information, only the monochrome camera may be used. The flow of the image quality determination method of this embodiment will be described below with reference to FIG.

Further, this embodiment will be described by using a transmission type color liquid crystal panel as the inspection object 5 in FIG. First
The image inputting means 11 which is the step of the step is the image of the transmissive color liquid crystal panel which is the inspection object 5 is digital data of R, G, B by the method shown in FIG. ) In the computer 7 of FIG.

Second step, data conversion means
In 12, the R, G, and B data are data that match the inspection means.
Convert to. In this embodiment, the R, G, and R on the RGB coordinates are
Convert B data to Y, x, y data on Yxy coordinates
You. In addition, the data conversion method is
RGB output value of camera 6 and color difference meter not shown
The Yxy output value of is calibrated in advance, and Yx is calculated from the RGB coordinates.
Creating a conversion table that converts data to y-coordinates
To convert R, G, B data to Y, x, y data
ing. In the data conversion means, Yxy
Coordinate is L^* a^* b ^* To convert to coordinates, conversion formula (1) X = Y (Y * (x / y) / y) * Y Y = Y (1) Z = ((1-x-y) / y) * Yxn = 100 yn = 95.045 zn = 108.892 L = 116 × (Y / yn)^1/3−16 a = 500 × ((X / xn)^1/3-(Y / yn)^1/3) B = 200 × ((Y / yn)^1/3-(Z / zn)^1/3) Is used. With the above conversion means, a transmissive color liquid crystal display
Nell Y, x, y data, L^* , A^* , B^* The data is
It is saved in the computer.

Next, in the inspection means which is the third step, Y, x, y, L ^* , a ^* , b ^* data obtained by the data conversion means 12 in the second step, or the first step.
The R, G, B data obtained by the image input means of the step is used as it is, and the inspection corresponding to the inspection item of the inspector is performed.

In the present embodiment, the image quality is evaluated by using four inspection methods as the inspection means: a luminance inspection, a chromaticity inspection, a color difference inspection, and an unevenness inspection. First, in the luminance inspection which is the first inspection method, the Y data which is the luminance data is used, and the inspection is performed by comparing the luminance standard value with all the luminance data. As shown in FIG. 3, the method of scoring the inspection result is to set the maximum allowable luminance value Y _max and the minimum allowable value Y _min , which are the standard values of luminance, to the maximum and minimum values of rank 5, and to set this standard value to and original, dY than the judgment formula (2) = (Ymax -Ymin) / 10 (2) Ymax 4 = Ymax -dY Ymax3 = Ymax4-dY Ymax2 = Ymax3-dY Ymax1 = Ymax2-dY Ymin4 = Ymin + dY Ymin3 = Ymin4 + dY Ymin2 = Ymin3 + dY Ymin1 = Ymin2 + dY The value dY is obtained by dividing the difference between Ymax and Ymin into 10 equal parts,
Also, the maximum value Ymax1 and the minimum value Ymi of rank 1 in FIG.
n1, rank 2 maximum value Ymax2, minimum value Ymin
2, maximum value Ymax3 of rank 3, minimum value Ymin3,
The maximum value Ymax4 and the minimum value Ymin4 of rank 4 are obtained respectively.

As shown in FIG. 3, rank 1 has one point,
Rank 2 is 0.8 points, rank 3 is 0.6 points, rank 4 is 0.4 points, rank 5 is 0.2 points, and ranks outside the range of rank 5 are ranks 6 and 0 points. Here, in the luminance inspection, it is investigated which of the ranks the Y data of the inspection image is included in, and the score of the included rank is added as the score for the Y data. As a result, individual scores are obtained for all Y data, and in this embodiment, the average of the scores is used as the score of the brightness inspection of the main inspection image. Here, since it is necessary to give importance to the result belonging to the rank 6 which is a value outside the standard value, the allowable number of data belonging to the rank 6 is set in advance, and when the allowable number is exceeded,
A process for setting the score to 0 is added.

Next, in the chromaticity inspection which is the second inspection method, the color difference inspection which is the third inspection method, and the unevenness inspection which is the fourth inspection method, as with the luminance inspection, Six ranks are divided based on the standardized value in advance, the rank and each inspection data are compared and inspected, and a score is set in the range of 0 to 1, and the result is output as the score of each inspection. Here, the scores obtained by the four inspection methods are v ₁ , v ₂ , v ₃ , and v ₄ .

The overall judgment means 14, which is the fourth step in FIG. 1, will be described below with reference to FIGS. 4 and 5. FIG. 4 is an image-based judgment table displayed on the monitor 8 in FIG. 2. The score results v ₁ , v ₂ ,
v ₃ and v ₄ are automatically substituted after the execution of the inspection means.
Also, the weights w ₁ , w ₂ , w ₃ , w ₄ for each inspection method
Is set in advance so that the judgment result and the subjective judgment of the inspector are associated with each other.

Further, based on each score and each weight, the number of judgment formulas is 1

[Equation 1] Thus, the image-based score V for the inspection image is obtained.

Here, since the transmission type color liquid crystal panel which is the inspection object of this embodiment is constructed so that full color display can be performed by using three elements of R, G and B as one pixel, in this embodiment An R display image in which only the R element is driven, a G display image in which only the G element is similarly driven, a B display image in which only the B element is driven, and a white (hereinafter referred to as W) display image in which all the R, G, and B elements are driven, Further, an inspection is performed on five types of display images of the gray (hereinafter referred to as Gr) display image in which the drive voltage for the W display is weakened.

Therefore, by performing the first to fourth steps in FIG. 1 for the five types of display images, the image-specific scores V ₁ , V ₂ , V ₃ , V ₄ , V ₅ of each display image are obtained. obtain.
These are substituted into the comprehensive judgment table shown in FIG. Also here, the weights W ₁ and W are assigned to the scores for each image so as to correspond to the subjective judgment of the inspector.
₂ , W ₃ , W ₄ , and W ₅ are set and are substituted into the determination formula 3 in the same manner as the method of obtaining the image-specific score V for each inspection image, and the total score S for the inspection object is calculated. Here, according to the method of calculating the score for each image, the score v of each examination is represented in the range of 0 to 1 as described above, and the weight w for each examination is also set in the range of 0 to 1, so the determination The image-specific score V calculated by the expression 1 is represented in the range of 0 to 1. In addition, in the present embodiment, the score V of each inspection described above is
Similar to the case of calculating, since the score 0, which means that the data outside the allowable value was included, is emphasized, it is investigated whether 0 is included in the score v of each inspection, and it is included. Does not use the judgment formula number 1, but adds processing for setting the score for each image to 0 point.

Also, when the total score S for the inspection object is calculated, the weight W for the image-specific score V is also set in the range of 0 to 1, and the total score S is calculated by the judgment formula number 1. It is expressed in the range of 0 to 1. Further, in the present embodiment, similarly to the above-described score calculation method, since the score 0, which means that the out-of-tolerance data is included, is emphasized, it is investigated whether the image-specific score V includes 0 points. However, if it is included, the process of setting the total score to 0 is added without using the judgment formula number 1.

By the above method, the judgment result for the inspection object can be obtained. Here, when it is desired to obtain a score as a determination result, the score calculated by the above-described method can be used as it is. Further, when it is desired to obtain the determination result by determining whether the product is a good product or a defective product, a threshold value is set between 0 and 1, and if the score is lower than the threshold value, the product is defective or higher than the threshold value. It can be judged as a good product.

Although the present invention has been described above with the transmission type color liquid crystal panel as the inspection object, the same determination method can be applied to any display device such as a monitor, a plasma display and an LED display. Further, any material can be applied as long as it evaluates and judges the image quality and unevenness of printed matter, coated surface of industrial products and the like, textiles and the like.

Since the judgment formula number 1 is a normalization formula,
The score can be calculated without depending on the number of substituted values. For example, in the present embodiment, it is possible to make a determination by the same determination method even if the inspection content and the number of inspections are changed for each image and the inspection is performed. However, in the present invention, even if the judgment formula is other than the expression 1, for example, all the scores,
It is possible to make a judgment even with a judgment formula that multiplies weights, and add the points multiplied by weights,
The determination can be made even by the method of taking an average, and the determination formula is not limited. Further, it is not necessary to limit the range of the standard value of the score and the weight from 0 to 1.

[0028]

The image quality judging method of the present invention has means for comprehensively judging from a plurality of inspection results, so that (1) it is possible to correspond to an inspection item which has been conventionally judged by an inspector. In addition, since a plurality of types of inspections are performed by the computer and the results thereof are scored within a standardized range, it is possible to treat different types of inspection results as values of the same dimension.

(2) Since it is possible to individually weight a plurality of inspection results, it is possible to make adjustments so that the judgment is closer to the subjectivity of the inspector. (3) By setting the standard value of the score and the weight to be 0 to 1 and calculating the score using the judgment formula number 1, the score obtained is also in the range of 0 to 1. Therefore, it is possible to handle the results obtained by performing the processing of the present method on different inspection images in the same dimension.

(4) Since the number of judgment formula number 1 does not depend on the number of substituted values and the score is calculated, for example, in this embodiment, the inspection contents and the number of inspections are changed for each image to perform the inspection. However, there is an effect that it is possible to make a determination by the same determination method.

[Brief description of drawings]

FIG. 1 is an operation explanatory diagram illustrating a flow of an image quality determination method of the present invention.

FIG. 2 is a configuration diagram showing an embodiment of the present invention.

FIG. 3 is an explanatory view of a ranking method for explaining a ranking method and a score adding method in the embodiment.

FIG. 4 is an explanatory diagram of an image-by-image determination method for explaining an image-by-image determination table in the embodiment.

FIG. 5 is an explanatory diagram of a comprehensive determination method for explaining a comprehensive determination table in the embodiment.

[Explanation of symbols]

 1 Transmission Lamp 2 Document Plate 3 Reflection Lamp 4 Dark Room 5 Inspection Object 6 Color Camera 7 Computer 8 Monitor 11 Image Input Means 12 Data Conversion Means 13 Inspection Means 14 Comprehensive Judgment Means

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Shin Ishiuchi 1-8 Nakase, Nakahama, Mihama-ku, Chiba, Chiba Prefecture SII AIRD Center, Inc. (72) Inventor Atsushi Uemoto Nakase, Mihama-ku, Chiba, Chiba Prefecture 8-chome, SII Co., Ltd.

Claims (4)

[Claims] 1. An image quality determination method for inspecting one image to be inspected by a plurality of inspection means, and determining image quality by a comprehensive determination means based on a plurality of inspection results obtained from the inspection means, It is possible to judge the image quality by expressing the score in a range in which the result of each inspection is arbitrarily specified by a plurality of inspection means, and calculating the score by the arbitrary calculation method by the total judgment means to obtain the total score. Characteristic image quality judgment method. 2. The image quality is set by setting a weight value for the plurality of inspection means in an arbitrarily standardized range, and calculating the score and the weight value by an arbitrary calculation method by the comprehensive judgment means to obtain a total score. The quality judgment method according to claim 1, wherein the quality judgment method is performed. 3. An inspection result obtained by inspecting a plurality of images by one or a plurality of inspecting means is expressed as a score in an arbitrarily standardized range, and the comprehensive judging means calculates the score by an arbitrary calculation method. An image quality determination method characterized in that the image quality of the plurality of images is determined by performing a calculation to obtain a total score. 4. The weight value for the plurality of images and the weight value for the plurality of inspection means are set within a standard range, and the score obtained by the inspection means and the weight value are calculated by an arbitrary calculation method. The image quality determination method according to claim 3, wherein the image quality is determined by obtaining a total score.