JP5176014B2 - Display multi-angle measuring system and method - Google Patents

Description

  The present invention relates to a multi-angle measurement system and method for quantifying unevenness on a display panel, and in particular, a data base for quality interactive processing analysis by taking images at a plurality of viewing angles on the panel by a multi-angle imaging mechanism. It means to make up.

  With the universal application of flat panel displays, for example, liquid crystal displays (LCDs) are used in large quantities on televisions, computer monitors, mobile phones, various home appliances, and the quality of liquid crystal display panels in mass production. For example, color, contrast, reaction time, brightness, etc. need to be more and more important, and regarding LCD background uniformity, that is, the generated unevenness due to LCD brightness abnormality is observed and unevenness is due to uneven display brightness. This refers to the various types of traces applied and display defects that can occur when a color filter is attached to the LCD. In addition, the driver chip (Drive IC) and the liquid crystal itself all cause unevenness.

  The simplest method of judging unevenness is to switch to the black screen and other halftone screens, and then observe in detail from various angles, and the liquid crystal display has various unevenness due to various process defects. For example, the defect may be a horizontal stripe pattern, a 45 degree stripe pattern, a square, or a circular mass, and a single mass may appear at a certain corner. Such a defect has a contrast, irregular display, and thus a serious unevenness. The degree is also an important reference for judging the quality of LCD panels.

  However, in the conventional technology, the judgment for unevenness is that a defect of the screen of the liquid crystal display is directly observed by an expert, and the degree of panel defect and the panel quality are judged based on the subjective knowledge. .

  1. It is difficult to detect and image.

  2. It is difficult to classify the unevenness of various situations, and the definitions of each eaves are not consistent, but the indicators are lacking.

  3. The standard of the serious degree of unevenness is not described or too poor.

  However, the international organization VESA (Video Electronics Standards Association) makes definitions for each type of irregularity (VESA FPDM2303-8). The degree of resolution and resolution of the judgment are not defined.

The conventional related technology partially proposes a nonuniformity detection apparatus and method, and among them, US6,154,561 proposes a nonuniformity classification, for example, the one shown in FIG. The line type unevenness of the flat panel display 10 is different from the vicinity and abnormal pixels, for example, straight line, curved line c, L type line a, vertical line b, fine line e, coarse line Unevenness such as f is generated. Further, for example, the block-type unevenness shown in FIG. 1B includes unevenness such as a dark spot block i and a bright spot block g, and further unevenness such as an etch block h and an etch bright spot block j generated at the panel boundary.
And, for example, US Pat. No. 5,917,935 proposes setting one threshold value to compare the mura defect with the background value. Therefore, the degree of the mura defect is obtained, a statistical table is created, and FIG. According to the flow shown and this statistical table, various different mura defect states are classified and defined. The step samples one primitive screen from the display panel (step 201) and generates a plurality of secondary sample screens from this primitive screen (step 202), based on different demands. Each secondary sample screen is filtered, for example, a specific feature filter is created based on various characteristic diagrams (histograms) of the screen (step 203), and one threshold value is set for each screen. Then, a feature block is generated (step 204). Further, the mura defect of the primitive screen is judged from the analysis of the feature block (step 205), and the mura defect is characterized (step 206). Then, the final adjustment operation is executed, and the error detection is deleted to determine a certain unevenness classification ( Includes a step 207), the.

  The conventional technology classifies only various types of irregularities, for example, line irregularities, spotted irregularities, filling irregularities, edge irregularities, block irregularities, etc., mainly by the steps of the above-mentioned patent. The quality of the display panel is not judged against the (pattern).

  The SEMI International Standard also announces the measurement definition (SEMI D31-1102) for screen quality irregularity inspection for flat panel displays. However, as long as any definition or analysis for unevenness takes into account the display area, contrast, and background brightness, it is too simple to completely exhibit flat panel display defects.

  The aforementioned conventional technique has the following drawbacks.

  1. Employs artificial detection but is therefore not very reliable and even controversial.

  2. Obtaining only mura physics descriptions or classifications does not take into account human perception and the model is too simple.

  3. Describe and classify only the severity of unevenness, but not the quality of the display.

  4. Many automatic detection techniques request segmentation detection, that is, too many parameters cannot be determined for image detection of various different unevenness, and once there is a new unevenness, Requires an arithmetic method.

  And the biggest problem of artificial detection is in variables such as fatigue due to human power and the degree of personal feelings, and it causes detection error or dropout phenomenon, so it greatly affects the consistency and stability of inspection quality, Therefore, there is a need to develop detection techniques that can be quickly quantified, thereby reducing decision errors and disputes.

  Conventionally, with respect to the unevenness phenomenon, it is limited to only identifying and defining various aspects, and if a detection unevenness is imaged using an image sensor device, some conventional systems are already provided. There is still a lack of judgment on the related human visual data base and unevenness diversity and changeability, technology is still immature, practical use is difficult, and consistency and quantification of quality control are difficult. In order to solve the generation of a judgment rule that is not objective and inconsistent using artificial detection in the prior art, that is, the present invention uses a multi-angle imaging mechanism to pick up the uneven phenomenon Furthermore, it is combined by the processing program of mura image analysis and mura identification classification, providing quantification mura detection technology, creating quality interactive processing analysis database, providing friendly interface and interactive search function .

  The multi-angle measurement system provided by the present invention includes one uneven image pickup unit, and uses one multi-angle image pickup mechanism to pick up an image on the display panel, and one uneven image processing unit. In response to the display panel image, the analysis, identification, and classification of the image is performed. Furthermore, there is one data base unit, and the interactive processing analysis data base is created by the data after the image processing. And provide an interface and an interactive search function.

  Among them, the multi-angle imaging mechanism includes a plurality of image sensor devices. However, a more preferable embodiment is that the image sensor device is implemented by a photosensitive element such as a CCD or a CMOS so that a plurality of viewing angles can be obtained. The display panel is photographed, and among them, a more preferable embodiment is a central image sensor device installed above the front viewing angle of the display panel, and an upper image sensor device installed above the vertical viewing angle of the display panel. Lower video sensor device installed below the vertical viewing angle of the display panel, left video sensor device installed to the left of the horizontal viewing angle of the display panel, and right installed to the right of the horizontal viewing angle of the display panel Includes image sensor devices. Each image sensor device (other than the central image sensor device) includes a photosensitive element implemented by connecting one oblique angle table to one camera / lens and one CCD or CMOS, so that the camera The main plane of the lens and the surface of the photosensitive element are inclined at an angle from each other.

  Another embodiment of the multi-angle imaging mechanism is an image sensor device installed on a track frame, and the image sensor device moves in a plurality of directions by a rail installed in the track frame. The display panel is photographed at the viewing angle, and the angle at which the main plane of the camera / lens included in the video sensor device and the surface of the photosensitive element are inclined with respect to the position of movement is changed.

  The steps of the multi-angle measurement method provided by the present invention are as follows. First, a single multi-angle imaging mechanism picks up images on a plurality of viewing angles on a panel, and then introduces a single human visual model. It generates two just-noticeable difference values, and the generation of just-noticeable difference values uses a human process method, performs qualitative and quantitative analysis on the unevenness, and the degree of visual detection of the human eye to the display panel. Create a database. Next, the unevenness of the display panel is judged and the unevenness is marked. Finally, a result of quantifying the unevenness aspect is generated, and the quantified result is stored in one data base.

  The above-described unevenness is determined by determining unevenness such as spots, lines, V-shaped striped patterns, scratches and light leakage. The generation of the just noticeable difference value uses a human process method, performs qualitative and quantitative analysis on the unevenness, and creates a database of the degree of visual detection of the human eye on the display panel.

  The present invention can provide a multi-angle measurement system and method for quantifying unevenness on a display panel.

It is a schematic diagram according to the classification of the nonuniformity in the nonuniformity measuring method of the prior art. It is a schematic diagram according to the classification of the nonuniformity in the nonuniformity measuring method of the prior art. It is a flowchart of the nonuniformity measuring method of the prior art. It is a schematic diagram of the 1st Example of the multi-angle imaging mechanism of this invention. It is a schematic diagram of the 2nd Example of the multi-angle imaging mechanism of this invention. It is a schematic diagram of the inclination angle between a video sensor apparatus and a panel. It is a schematic diagram of angle correction of the internal photosensitive element of the image sensor device of the present invention. The flow of the trapezoid correction of this invention is shown. 3 shows the flow of uneven image processing software of the present invention. It shows the flow of calculation of the just noticeable difference of the present invention. It is a schematic diagram showing construction of the interactive processing analysis data base of the present invention. It is a flow of the operation | movement step of the nonuniformity detection software of this invention.

  The mura phenomenon is diverse and changeable and cannot be detected entirely by a single viewing angle, i.e., the present invention provides one or more high resolution charge coupled device (CCD) video cameras or complementary metal oxides. Using a film semiconductor (CMOS), the display panel is imaged and detected from a plurality of different angles, and the multi-angle measurement system (1) picks up the uneven phenomenon using one multi-angle imaging mechanism. (2) A mura video processing unit that performs mura video analysis and mura identification classification processing by a single mura video processing program to provide quantified mura detection technology, and (3) quality interactive A database that provides a friendly interface and interactive search functionality by creating a processing analysis database And a unit.

  The unevenness phenomenon is diverse and changeable, and not all can be detected by a single viewing angle. Therefore, the main element of the unevenness image pickup unit of the present invention includes a single multi-angle imaging mechanism, which May include picking up images of different viewing angles and turning each angle of 25 degrees or other suitable angle vertically in addition to the front viewing angle, each 45 degree angle or other suitable angle Can be detected by turning it horizontally.

  A more preferred embodiment of the present invention is, for example, a schematic diagram of a multi-angle imaging mechanism shown in FIG. 3A, which operates five image sensor devices (31, 32, 33, 33, 34) and shoots an object. Among them, a more preferred embodiment implements the image sensor device with a photosensitive element such as a CCD or a CMOS, for example, in the present invention, the unevenness phenomenon is detected by imaging, but the mechanism faces the panel 30 or Central video sensor device 31 on the front viewing angle of the object, display panel or upper video sensor device 32 installed above the vertical viewing angle of the object, lower video sensor installed on the display panel or below the vertical viewing angle of the object Device 33, display panel or left image sensor device 34 installed to the left on the horizontal viewing angle of the object, and display Over panel or a right image sensor device 35 which is installed to the right on the horizontal viewing angle of the object. A multi-angle measurement system of a large number of video sensor devices shown in FIG. 3A can pick up images of a plurality of viewing angles of the panel 30 or the object, and generate a non-uniformity aspect based on a more precise judgment criterion. An image obtained by photographing two or more image sensor devices at different angles when actually executed is a main invention, and is not limited to the five image sensor devices shown in this embodiment.

  In order to achieve the imaging purpose of the multi-angle measurement system of the present invention, FIG. 3B shows another more preferable embodiment, in which a certain image sensor device 36 is installed on a track frame 39, and the image sensor device. The rail 36 installed in the track frame 39 can move in the horizontal, vertical or other directions. As shown in the drawing, the video sensor device 36 installed on the left side of the panel 30 The video of the left viewing angle is taken, but it can be moved to the position 36 ′ of the lower video sensor device using the track frame 39, the video of the lower viewing angle of the panel 30 is taken, or the video sensor device of the right side is taken. Moved to position 36 ", thereby taking a picture of the right viewing angle of the panel 30, or others at any position on the track frame 39, In this way, an image at a specific viewing angle of the panel 30 is taken in this manner, so that only one image sensor device is used, or an image sensor device installed in the center (not shown in the drawing) is further added. By increasing the number by one, the purpose of the multi-angle photographing panel 30 can be achieved, and among them, the main plane of the camera / lens and the surface of the photosensitive element of each image sensor device are inclined with respect to each other according to the position of movement. Change, and correct the depth of shooting.

  In the above-described multi-angle imaging mechanism, that is, the image sensor device positioned on the vertical viewing angle and the horizontal viewing angle (or other viewing angle) needs to overcome the measurement error problem caused by imaging the depth by the tilt angle. If the image is taken from different viewing angles, the depth imaged from the tilted angle will be insufficient. It was generated when the image sensor device irradiates the panel as shown in FIG. 4A, for example, as shown in FIG. 4A.

  That is, the present invention uses an optical mechanism imaged at an inclined angle to solve an error caused by the inclined angle. For example, in the embodiment shown in FIG. Images are taken from the right side of the panel 30, and both ends (end point a and end point b) of the range over which the image is taken have different degrees of image forming states. In the present invention, one oblique angle table is designed via optics and can be connected to a camera / lens and an image sensor device, respectively, and the angle between the surface of the image photosensitive element and the image forming surface is changed by this optical mechanism. 4 corrects and solves the problem that the depth imaged from the inclination angle is insufficient, and the image imaged from the oblique angle is clear, for example, a schematic diagram of the angle calibration of the internal photosensitive element of the image sensor device shown in FIG. 4B. As described above, by connecting one oblique angle table 47 to the main plane of the camera lens 41 of the right image sensor device 35 by optical design, the light beam transmitted through the camera lens 41 and the surface of the image photosensitive element 43 are A certain angle can be inclined with respect to each other, thereby extending the depth and achieving the imaged purpose with matching resolution.

  In a more preferred embodiment, for example, from the both ends a and b of the photographing range of the panel 30 in the drawing, the camera lens 41 is transmitted and focused on the camera lens aperture 45, and the image is again displayed through the oblique table 47. Projects onto the photosensitive element 43. Since each point picked up from the top of the panel 30 has the same resolution (the same depth), the image photosensitive element 43 and the camera lens 41 are inclined at an appropriate angle, and an image is captured. Each point on the panel 30 has the same depth. Therefore, shooting of various viewing angles on the various corresponding panels 30 requires a video sensor device having the ability to relatively tilt the video photosensitive element 43 at a certain angle.

  Particularly, for example, as shown in the schematic diagram of the configuration shown in FIG. 4B, it is necessary to form an image on the image photosensitive element 43 via the camera aperture 45 or other optical mechanism, so that the image photosensitive element is captured. The element 43 can be inclined at an angle based on an actual mode, and each point on the panel 30 can have the same depth when an image is formed on the image photosensitive element 43.

  The uneven image processing unit of the present invention further uses a uneven image processing program and combines detection techniques that can be quantified. Among them, for example, a method of combining human processes with an uneven image analysis and unevenness classification processing program Qualitative analysis, quantitative analysis, and experiments are performed to create a data base for the degree of visual detection of the human eye on the display panel. Expect consistency and quantification. The uneven image processing unit mainly includes the following.

  (1) The trapezoidal distortion correction program imaged at an inclined angle has a non-uniformity corresponding to an inclined image at an accurate position on the panel and contributes to the operation of the image processing program.

  (2) The mura image analysis and identification classification processing program can detect mura defect modes and includes spots, lines, V-shaped striped patterns, scratches and light leakage, but is not limited to this mode of description.

  (3) The mura image processing program combines a human vision (vision) model database of interactive processing analysis quantification of detection technology that can be quantified. In other words, the human process method is used, qualitative and quantitative analysis is performed for unevenness, a data base for the degree of visual detection of the human eye on the display panel is created, and this data base is used for the production of quality control work standards. As a basis, we expect further consistency and quantification of quality control. Its technical goal is to provide an objective and stable method that can be applied to production lines, quantifying uneven defects, and combining panel images acquired by an imaging system. ”Is used to calculate the psychological strength of the mura defect of the entire panel and display it with just noticeable difference (JND).

  (4) The interactive processing analysis data base forms an unevenness quality interactive processing analysis data base by the interactive processing analysis data base, and provides a friendly interface and an interactive search function.

  The means used for the keystone correction in the above description includes, for example, as shown in FIG. 5, firstly defining coordinates in the trapezoidal image (step S501), but after being imaged by the multi-angle measuring device described above. The software picks up the position of each pixel in the image and defines the position of the coordinate, and then converts the coordinate (step S503), that is, by the photosensitive element executed by the CCD or CMOS by the software. If the picked-up pixel position is converted into various coordinate numerical values, and then the corresponding calculation of the pixel and coordinates is performed (step S505), the trapezoidal distortion state can be determined. Correction is performed (step S507), and the problem of image distortion caused by photographing at an inclined angle is calibrated.

  FIG. 6 shows a main software flow of the uneven image processing of the present invention, and a quantified human visual data base is generated by interactive processing analysis of the quantified detection technique. At the beginning, after taking the measurement waiting panel using the above-mentioned multi-angle imaging mechanism, the imaging is performed (step S601), and after the imaging, pre-processing before analysis is performed (step S603), among which the rear end computer The system includes recording of image pixel information captured by each image sensor device, for example, information such as pixel coordinates, brightness, and imaging area, and also includes the trapezoidal correction step described above.

  Next, unevenness image analysis is performed, and unevenness analysis of various aspects is performed by the software method provided by the present invention. From the result of analysis, unevenness in the panel, such as spots, lines, V-shaped stripes, scratches, Light leakage or the like is determined (step S605). This flow can identify, analyze, and quantify one or more types or types of unevenness, and if the unevenness in the panel is to be processed, the operational analysis of the unevenness of the spot is performed. The spot unevenness is acquired and marked (labeled), that is, a video message such as the coordinates, brightness, and area of the spot unevenness is marked (step S607).

  Quantification is performed from the above-described analysis result (step S609), and for example, a contrast value of unevenness is calculated and corrected by adding a human parameter, for example, just a notable difference value (JND value). A data base is generated, an objective unevenness judgment is given by an actual number, and a report table is generated by this number or stored in the data base (step S611). Other various aspects of irregularities also need to be quantified through the steps of identification, analysis, etc. described above and to generate an interactive processing analysis database of human vision parameters introduced according to the present invention.

  The above-described calculation of the just noticeable difference value is, for example, the flow shown in FIG.

  First, one reference image is introduced (step S701), and the unevenness image generated by the above-described analysis flow is introduced (step S703). After that, the reference image and the uneven image are compared with each other (step S705), and further the contrast sensitivity filter calculation (CSF) is performed (step S707), thereby calculating the visual difference of the human eye (step S709). After this, the average value of the difference of each pixel in the video is calculated, and a more preferred embodiment among them is calculated by the Minkowski pooling average value method (step S711), and thus just the noticeable difference (JND). Is calculated (step S713).

  The operation of the above-mentioned interactive processing analysis data base refers to the data base construction diagram shown in FIG. In order to create one uneven quality interactive processing analysis data base, this data base must include a data creation mode and a data display mode. As shown in FIG. It includes an update module 81 and a data display mode 83, and it is necessary to create this data base before using the data, and includes functions such as a new data increase 811, a data correction 813, and a data deletion 815. In order to contribute to the subsequent determination of unevenness, and to introduce a panel specification under the data display mode (831), to identify unevenness, and to identify unevenness A result is generated (833), and the result after further identification is displayed by a report table (835). Ri by providing detection techniques quantification unevenness, created a quality interactive process analysis data base, provides the functions of friendship interface and interactive search.

  The present invention is a multi-angle automated imaging device that includes a number of image sensor devices (photosensitive elements such as CCD or CMOS) to capture images at a plurality of different viewing angles, and includes an image photosensitive element to provide an inclination angle. By generating, the depth optical configuration can be calibrated at the time of imaging, but there is core unevenness detection software. Among them, the unevenness image processing program includes a trapezoidal distortion correction program imaged at an inclined angle, unevenness image analysis and unevenness An object of the present invention is to combine a processing program for identification classification and a visual model corresponding to determination of human vision and to give a quantification, combined with a detection numeral interactive processing data base, etc. To achieve.

  The flow shown in FIG. 9 is an operation step of the unevenness detection software of the present invention, among which an uneven image pickup step (uneven image pickup unit), an uneven image processing step (uneven image processing unit) and an interactive processing analysis data base. (Data base unit) Three parts such as search are included.

  At the beginning, it is necessary to install and end the apparatus related to the multi-angle measurement system provided by the present invention, for example, install a measurement waiting panel, install a multi-angle imaging mechanism, and further link a data base, etc. Complete the job.

  Next, panel image pickup at a plurality of viewing angles is performed by a multi-angle imaging mechanism (step S901), and a more preferable embodiment of the panel image pickup step is five image sensor devices (31, 31, 31) shown in FIG. 32, 33, 33, 34) to capture unevenness, and other embodiments, such as a video sensor device on five viewing angles corresponding to the panel shown in the figure, are shown in FIG. Photographing in each viewing angle direction is performed by operating one image sensor device installed on the frame. After picking up images from a plurality of viewing angles, a human visual model is introduced on the one hand (step S903), and a JND value is generated (step S905). Including a base and providing an interface and interactive search functionality, so that the data-based interface contrasts the picked-up panel video with the data in the data base, thereby reducing the diversity aspect on the panel. After determining a specific unevenness (step S907), a step of marking (labeling) the unevenness is performed (step S909). The specific unevenness of the description is determined by repeating the determination step to determine unevenness such as spots, lines, V-shaped striped patterns, rubbing traces, and light leakage simultaneously, and is not limited to the exemplary embodiment here. .

  After generating the above-mentioned JND value and judging the mode and the marking of the unevenness, that is, the human visual model operates the human process method, performs qualitative, quantitative analysis and experiment on the unevenness, A database for the degree of eye vision detection is created and quantified results are generated (step S911), and one of them is the mura image processing program, which can quantify detection technology interactive processing analysis quantification of human vision By combining the model data, a quality interactive processing analysis data base is created, and a friendly interface and interactive search function are provided (step S913). Finally, this is used as the basis for quality control work standard production, and the purpose of consistency and quantification measurement is achieved.

  The foregoing accompanying drawings are provided for reference and explanation only and are not used to limit the present invention.

  Summarizing the above description, the present invention applies an optical configuration of increased depth of bevel calibration and a trapezoidal distortion correction program to restore an image, and at the same time, uses a “human visual model” to detect mura defects on a panel. We provide an objective and stable measurement system that can be applied to production lines by creating a quantification, adding a processing program for image analysis and identification classification of unevenness, and creating a quality interactive processing analysis database.

  The foregoing description is merely a preferred and feasible embodiment of the present invention, and is not intended to limit the scope of the present invention, and is therefore equivalent by operating the contents of the present specification and drawings. Any change in configuration is synonymously included in the scope of the present invention.

a L-shaped line b Vertical line c Curve e Fine line f Coarse line g Spotlight area h Etch area i Dark spot area j Etch spotlight area 10 Flat panel display 30 Panel 31 Central image sensor device 32 Upper image sensor device 33 Lower image sensor device 34 Left image sensor device 35 Right image sensor device 36 Image sensor device 36 ', 36 "Position of image sensor device 39 Track frame 41 Camera lens 43 Image photosensitive element 45 Camera aperture 47 Angle table

Claims (21)

A multi-angle measuring system for display,
An uneven imaging unit that uses one multi-angle imaging mechanism to pick up images on one display panel;
A non-uniformity video processing unit that receives video from the display panel and performs video analysis, identification, and classification processing;
A data base unit that creates an interactive processing analysis data base from the data after the video processing and provides an interface and an interactive search function;
Including
The multi-angle imaging mechanism includes an image sensor device installed on one track frame, and the image sensor device moves in a plurality of directions by a rail installed in the track frame, Take a picture of the display panel according to the viewing angle,
The image sensor device includes one camera / lens and one image photosensitive element, and among them, the main plane of the camera / lens and the surface of the photosensitive element are inclined to each other at a predetermined angle according to the position of movement. ,
The unevenness image processing unit is used for detecting a trapezoidal distortion correction program for correcting trapezoidal distortion for an image on the display panel after being imaged at the tilted viewing angle, and for detecting a plurality of unevenness defects. For the detected unevenness combined with the processing program of mura image analysis and identification classification including aspects such as spots, lines, V-shaped stripes, scratches and light leakage, and human visual model database that can be quantified A mura and video processing program that performs qualitative and quantitative analysis and creates a database of the degree of visual detection of the human eye for the display panel, and an interactive processing analysis data base including a single interface and interactive search function, Murata's quality interactive processing analysis database Look,
The human visual model is displayed by just one notable difference value (JND).
A multi-angle measuring system for displays.   2. The multi-angle measurement system for a display according to claim 1, wherein the main plane of the camera lens and the surface of the photosensitive element exhibit an inclination of the angle by one oblique angle table.   2. The display multi-angle measuring system according to claim 1, wherein the image sensor device is implemented by a single CCD photosensitive element.   2. The display multi-angle measuring system according to claim 1, wherein the image sensor device is implemented by a single CMOS photosensitive element. The trapezoidal distortion correction program means is:
Definition video coordinate means for defining the coordinates of the trapezoidal video,
Coordinate conversion means for converting the position of the image pixel into a coordinate value;
A coordinate correspondence calculation means used to correspond to the coordinates and the pixels;
Insertion means for calibrating the strain image;
The multi-angle measuring system for a display according to claim 1, comprising: Means for the processing program of the uneven image analysis and identification classification are as follows:
Unevenness analysis means for analyzing the photographed unevenness aspect;
Mura marking means for marking the video message of the Mura video,
Unevenness quantification means for generating a quantified value of the unevenness image,
Data base storage means for storing the quantified value in one human visual data base;
The multi-angle measuring system for a display according to claim 1, comprising:   7. The display multi-angle measuring system according to claim 6, wherein the unevenness marking means includes marking the coordinates, luminance and area of a plurality of unevennesses.   7. The display multi-angle measuring system according to claim 6, wherein the unevenness quantification means includes generating the human visual data base by matching the contrast value of the unevenness with a human parameter. The means for calculating the just noticeable difference value is:
Reference video introduction means to introduce one reference video,
Mura video introduction means for introducing the Mura video,
Video contrast comparing means for comparing the reference video and the unevenness video;
A contrast sensitivity filter operation means for performing a contrast sensitivity filter operation on the reference image and the unevenness image;
Visual difference calculation means for the human eye that calculates the visual difference of the human eye by the video contrast comparison means and the contrast sensitivity filter calculation means;
An average value calculating means for calculating an average value of differences of each pixel in the video;
Just-noticeable difference calculation means for calculating just-noticeable difference by the average value,
The multi-angle measuring system for a display according to claim 1, comprising:   10. The display multi-angle measurement system according to claim 9, wherein the average value calculation means uses a Minkowski pooling average value method. The interactive processing analysis database is
A data update module including one data new increase means, one data correction means and data deletion means;
A data display module including one panel introducing means, one unevenness identifying means and one result reporting means;
The multi-angle measuring system for a display according to claim 1, comprising: A multi-angle measurement method for a display used to measure and quantify unevenness on a display panel,
Picking up images on the display panel, and picking up images on a plurality of viewing angles on the panel by a multi-angle imaging mechanism;
Introducing a human visual model,
Generating exactly one notable difference value;
Determining the unevenness of the display panel;
A step of marking the unevenness;
Quantifying the unevenness aspect;
Storing the quantified results in one database;
Including
The multi-angle imaging mechanism includes a video sensor device installed on one track frame,
The step of picking up the image of the display panel is to move the image sensor device in a plurality of directions by means of rails installed in the track frame, and to photograph the panel of the display from a plurality of viewing angles.
The video sensor device includes one camera / lens and one video photosensitive element, and the main plane of the camera / lens and the surface of the photosensitive element are inclined to each other at a predetermined angle according to the position of movement,
The generation of the just noticeable difference value uses a human process method, performs qualitative and quantitative analysis on the unevenness, and creates a database of the degree of visual detection of the human eye for the display panel.
A multi-angle measurement method for displays.   13. The multi-angle measurement method for a display according to claim 12, wherein the main plane of the camera lens and the surface of the photosensitive element exhibit an inclination of the angle by one oblique angle table.   13. The display multi-angle measuring method according to claim 12, wherein the image sensor device is implemented by a single CCD photosensitive element.   13. The display multi-angle measuring method according to claim 12, wherein the image sensor device is implemented by a single CMOS photosensitive element.   13. The display multi-angle measuring method according to claim 12, wherein the unevenness is determined by determining unevenness such as spots, lines, V-shaped stripes, scratches and light leakage.   The step of picking up the image of the display panel further includes a non-uniformity image processing step, and the non-uniformity image processing step includes correcting the trapezoidal distortion, and correcting the trapezoidal distortion of the panel image captured at the tilted viewing angle. 13. The display multi-angle measuring method according to claim 12, wherein: The trapezoidal distortion correction step is as follows:
Defining the coordinates of the video,
Converting the video coordinates;
Calculating means of the video coordinates;
Inserting the video coordinates,
The display multi-angle measuring method according to claim 17, further comprising:   13. The display multi-angle measuring method according to claim 12, wherein the step of marking the unevenness includes marking a plurality of unevenness coordinates, luminance and area. The step of calculating the just noticeable difference value is
Introducing one reference video,
Introducing the unevenness image in the display panel,
Comparing the reference image with the unevenness image on the display panel;
Performing control sensitivity filter calculations;
Performing visual difference calculations for the human eye,
Calculating the just noticeable difference by calculating one average value;
The multi-angle measuring method for a display according to claim 12, comprising:   21. The display multi-angle measuring method according to claim 20, wherein the average value calculating step uses a Minkowski pooling average value method.

Images