JPH06308036A - LCD image inspection method - Google Patents

Abstract

PURPOSE:To reduce data amount without degrading the image data and inspect with high speed and high accuracy by taking in a normal display image of an element to be inspected in advance and comparing with the displayed image of tested body and judging between normal/abnormal. CONSTITUTION:The image data taken with a TV camera 1 is contained in a memory 2 and the contained image or analytical result is displayed on a monitor 3. A tested device controller 5 controls the indication or motion of a tested device 4 and a computer 6 controls the image memory 2 and the controller 5 and also analyzes the image data. By taking in a normal display image of an element to be inspected in advance which is compared with the displayed image of a tested body, normal/abnormal of an LCD indication is judged. Only by data formed on the basis of outline of the element to be inspected or limited specific region in the element not by whole region of the element, the indication existence is checked and thus, data processing is simplified and expedited.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a display image inspection of an apparatus equipped with a display element such as a liquid crystal display, and more particularly to an efficient method for realizing visual inspection automation in a mass production line of the apparatus. Liquid crystal display (Liquid Cry)
Since a stall device (hereinafter referred to as LCD) can easily display not only numbers but also a special pattern according to the purpose, and has features such as low power consumption, thin shape, and light weight, it has been rapidly popularized in recent years. However, the LCD has a weak point that it is more susceptible to high heat than general circuit parts, and it is not preferable to mount it on a printed circuit board together with other parts and pass it through a solder bath. For this reason, the LCD and the printed circuit board can be assembled manually by a process different from the solder dip on the printed circuit board assembly line, such as crimping with a rubber connector, connection with a flexible printed board, or soldering with special pins. Many. Because of this, the incidence of connection failure due to the structure and work is higher than that of other parts, and it is necessary to perform display inspection on all the produced products in terms of device manufacturing.

[0002]

2. Description of the Related Art In a conventional display image inspection of a liquid crystal display or the like, a drive signal applied to a terminal of a display element is electrically inspected or a display content of the display element is visually confirmed by a person.

[0003]

For this reason, the former has a problem that the automatic inspection is possible, but the connection failure between the drive circuit and the display element cannot be inspected, and the latter has a long inspection time.

[0004]

SUMMARY OF THE INVENTION The present invention provides an inspection method by which an image processing technique is applied to enable a high-speed and high-precision display image inspection without visual inspection. The present invention enables high-speed and high-precision display image inspection by reducing the data amount according to the characteristics of the display image without deteriorating the quality of image data to be inspected.

The inspection method of the present invention will now be described in detail with reference to the drawings. FIG. 1 shows an embodiment of the inspection system configuration according to the present invention, in which 1 is a television camera for storing images, and 2 is a television camera. Is an image memory for storing image data, 3
Is a monitor for displaying stored images or analysis results, 4 is a device under test having a display image to be inspected, 5 is a device under test control unit for controlling display or operation of the device under test, 6 is an image memory and a device under test 3 is a block diagram showing a computer that controls a device control unit and analyzes image data. [Basic Concept of Inspection] The basic concept of the inspection according to the present invention is that a normal display image of an element to be inspected is captured in advance and collated with the display image of the DUT to thereby display the LCD.
It is to judge whether the display is normal or abnormal. For example, when the specific element is not displayed due to the contact failure of the connector, it is determined as "abnormal" because it is different from "normal value (displayed)".

In general, an LCD is applied with a voltage between electrodes arranged inside thereof to rotate the polarization plane of the light by applying a voltage to the polarization of the light, in combination with a polarizer placed on the surface of the LCD. It controls transmission / blocking of light and displays "white" or "black" on the screen. The display pattern is determined by the shape of the electrode, and the same electrode (pattern)
Since the display density is extremely uniform inside, the presence or absence of display need only be inspected for a part of the electrode (pattern) rather than the entire area. The inspection method of the present invention pays attention to this point, and checks the presence or absence of the display in the data created based on the outline of the element or only in the limited specific area within the element, not in the entire area of the arbitrary element to be inspected. By doing so, the data processing method is simplified and the processing speed is increased.

Next, seven bar-shaped display elements (segments)
The inspection of the liquid crystal display element that displays a one-digit number as an image will be described. [Storage of Image Data] FIG. 2 shows an example (1200) of a display image of a 4-digit 7-segment number, and the image density is detected along an arbitrary scanning line a. In this case, when the non-display image data and the image data in the state in which the segment to be inspected is displayed are stored and the density difference between the two data is obtained for the same coordinates, the difference value of the segment is compared with the background The difference value becomes smaller. That is, as shown in FIG. 3, an example of density difference data in which the result obtained by the above-mentioned processing for the scanning line a shown in FIG. 2 is shown by taking the scanning position on the X axis and the density of the display image on the Y axis. so,
The peak value of the higher density is shown as the segment where the segment is displayed, and the low level is shown as the background.

[Detection of Segment Edge (Boundary Line)] In FIG. 3, for example, the density (25) is used as a threshold value, and the X coordinate point that transitions from a value smaller than the threshold value to a larger value is the start point edge, and from the value larger than the threshold value. X transitioning to a smaller value
A starting point edge and an ending point passing through the threshold line are detected with the coordinate point as the ending edge. By performing the same processing on all the scanning lines for scanning the image of FIG. 2 and arranging them on a two-dimensional plane corresponding to the stored images, data of substantially segment contours can be obtained. An example of edge detection obtained by the above method when "8" is displayed as a one-digit seven-segment number is shown in FIG.

[Segment normalization] In FIG. 4, if a series of midpoint data in which the midpoints from the start edge to the end edge corresponding to each segment are calculated is used, 1
Each segment can be expressed by a set of points, and “segment normalization” can be performed. FIG. 5 shows an example of segment normalization obtained by applying the above-described method to the edge detection example shown in FIG.

[Data Division] Next, the standardized data shown in FIG. 5 is divided into five lines (a, b, c,
Segments are divided according to d and e), it is checked which segment each midpoint data belongs to, and the number of data belonging to each segment is obtained.

[Display Judgment] Next, when the number of midpoint data obtained for each segment matches or exceeds a certain threshold value, it is determined that the segment is displayed, and the threshold value is exceeded. If not, it is determined that it is not displayed.

[Inspection] By the processing method described so far,
A normal display image serving as an inspection standard is stored in advance as a reference image, and then stored as a display image of the device under test. After that, the difference value between the reference image represented by the data compressed by the above-described data processing method and the inspection image is obtained, and if the difference value matches or is less than a certain threshold value, the inspection image is “normal”. It is determined that the inspection image is “abnormal” when the mismatch or the difference exceeds a certain threshold.

[Coordinate conversion for each character]
In the image displaying segment numbers, the coordinate conversion in number units for separating each digit and enabling the display inspection in units of one digit will be described. 7 shows an image of the liquid crystal display element, and scanning is performed from the approximate center of the non-display image toward the end of the display image in the arrangement direction of the numerical display units in FIG. The point i is defined as the boundary reference position of the display. Next, as shown in FIG. 8, the image is divided into areas including a 1-digit number display unit from the boundary reference position at regular intervals and converted into 1-digit unit display data, and then the 1-digit display inspection is performed for each area. The method detects the presence or absence of a segment.

FIG. 9 shows a second embodiment of the present invention, which is a display example of a display consisting of a numeral consisting of a 7-segment display and a character or symbol pattern indicating the operating state of the apparatus. The inspection of the liquid crystal display image displaying an arbitrary pattern such as the numbers and letters or symbols of the 7-segment display will be described.

[Display Area Extraction] FIG. 10 shows a method for extracting a display area to be inspected from the image information of FIG. 9, and a straight line 1 is an arbitrary scanning line for scanning the image. The image showing the display in the non-display state shown in FIG. 10 is scanned from approximately the center of the displayed image toward the end along two inspection lines set in parallel with the scanning line 1 to suddenly change the density. The boundary positions (a, b, e and f) of the display are determined. Similarly, the boundary positions (c, d, g and h) are determined for the two inspection lines set perpendicular to the scanning line 1. Of these eight points, straight lines ab and c
The display image frame is calculated from the four straight lines obtained by the combination of the two points d, ef, and gh, and the display area is extracted with respect to the image data stored in the image memory.

[Setting of Coordinate Reference] In a display image displaying an arbitrary pattern, image data that can be always detected is set as the coordinate reference. The image data that can be constantly detected may be any display element of the display device, a print pattern on the display device, a corner portion of the display frame, or the like that can be fixedly captured as image data. In FIG. 9, a clock mark is always displayed for time notification, and its center can be used as a coordinate reference.

[Setting of Inspection Point] Next, a method of setting an inspection point will be described with reference to FIG. 11, in which the coordinate reference of the image obtained in the above process is used as the first inspection position reference.
A second inspection standard is defined within each pattern to be inspected. After that, an inspection area that is included in the inspection target pattern or includes a part of the inspection target pattern is defined on the basis of the second inspection reference, and the inspection target image and the standard image density are compared only for the inspection area, The presence or absence of the display pattern is detected. By this processing, the inspection data amount can be reduced without deteriorating the quality of the stored image data, and therefore the inspection time can be shortened to an extremely short time.

[Inclination Correction of Inspection Screen] In the inspection system shown in FIG.
The position of the stored image differs depending on the position of. In particular,
When the image rotates on a plane and the inclination with the scanning line becomes large,
Differences in the coordinate positions of the corresponding patterns between the reference image and the inspection image are difficult to make, making accurate comparison difficult. In order to eliminate this adverse effect, the present invention provides the following image tilt correction method. The tilt correction value of the display image with respect to the scanning line 1 is calculated from any of the four straight lines ab, cd, ef, and gh obtained in the display region extraction processing, and the tilt of the inspection target screen is corrected by the tilt correction value. To do.

The tilt correction can also be obtained by using a pattern that is always displayed on the display screen. For example, with respect to the clock mark shown in FIG. 9, the pattern is scanned along each of m horizontal inspection lines and vertical n inspection lines, and the position where the density changes abruptly is the boundary position (edge) of the display pattern. Then, the tilt correction value of the display pattern is calculated from the obtained 2 (m + n) boundary values, and the tilt of the inspection target screen is corrected. That is, when setting the inspection point, the inspection point can be correctly set by correcting the coordinate value using the inclination correction value, and the comparison of the corresponding patterns between the reference image and the inspection image can be accurately performed. it can.

Although the display image shown in FIG. 9 has been described above, the present invention can also provide the same effect for an image displaying numbers or characters by the combination of mosaic patterns shown in FIG.

[0021]

According to the present invention, by comparing the image information created from each of the reference display image and the display image to be inspected, a high-speed and high-precision display image inspection can be performed without visual inspection. It should be possible to compare by checking the presence or absence of the display image only in the data created based on the outline of the element or only in the limited specific area within the element, instead of the entire area of the arbitrary element to be inspected. It is possible to reduce the amount of data and to achieve a great effect of enabling high-speed and high-precision display image inspection without deteriorating the quality of image data to be inspected.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an inspection system showing an embodiment of the present invention.

FIG. 2 is a diagram showing an example of a display image of numbers including 7 segments.

FIG. 3 shows density difference data of 7-segment numbers.

FIG. 4 shows a display for edge detection of 7-segment numbers.

FIG. 5 is a diagram showing standardization of segments.

FIG. 6 is a diagram showing division of segments.

FIG. 7 shows how to obtain a character division line.

FIG. 8 shows coordinate conversion in character units.

FIG. 9 is a diagram showing an example of a display image of a clock mark on the LCD screen.

FIG. 10 shows a method of extracting a display area.

FIG. 11 shows setting of inspection points.

FIG. 12 shows an example of a display image on a matrix display.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Television camera 2 ... Image memory 3 ... Monitor 4 ... Inspected device 5 ... Device under test control unit 6 ... Computer

Claims (7)

[Claims] 1. A display screen to be a reference is stored as first image information and a display screen to be inspected is stored as second image information in a storage device, respectively, and then the first image information and the second image information are stored. The feature of the image is extracted from the image to reduce the amount of information, or the third information and the fourth information limited to an arbitrary limited area in the display element are detected, and the third information and the fourth information match. Alternatively, the second image information is determined to be normal when the density difference is less than or equal to a certain threshold value, and the second image information is determined to be abnormal when the mismatch or the density difference exceeds a certain threshold value. A method for inspecting a liquid crystal display image, which is characterized by: 2. In the inspection of a one-digit numerical display image composed of seven rod-shaped display elements (segments), an arbitrary segment is scanned in the horizontal direction to obtain the coordinates of the starting point and the ending point of the segment, and After calculating the intermediate data between the end points and creating one set of intermediate data for one segment,
The intermediate data is divided for each segment, it is determined that the segment is displayed when the sum of the intermediate data values shows a value equal to or more than a certain threshold value, and the intermediate data value is a value less than the certain threshold value. Is displayed, it is determined that the segment is not displayed. 3. A one-digit numerical display unit is composed of seven rod-shaped display elements (segments), and in the inspection of a display image composed of a plurality of the numerical display units, there is nothing along the arrangement direction of the numerical display units. The boundary reference position of the display image is defined as a position at which the density is rapidly changed by scanning from the approximate center of the display image to the edge of the display image, and the display image is divided from the boundary reference position into a region including a one-digit number display unit. After that, the liquid crystal display image inspection method is characterized by detecting the presence or absence of a segment for each area. 4. In a display image displaying an arbitrary pattern, m lines in the horizontal direction and n lines in the vertical direction on a non-display image.
Scanning is performed from the approximate center of the display image toward the end along each of the inspection lines of the book, the position at which the density suddenly changes is defined as the boundary position of the display image, and 2 (m + n) boundary positions obtained here are determined. A method for inspecting a liquid crystal display image, wherein a display image frame is calculated from data and the inside of the display image frame is set as an inspection target area. 5. In a display image displaying an arbitrary pattern, m lines in the horizontal direction and n lines in the vertical direction on the non-display image.
Scanning is performed from the approximate center of the display image toward the end along each of the inspection lines of the book, the position at which the density suddenly changes is defined as the boundary position of the display image, and 2 (m + n) boundary positions obtained here are determined. A method for inspecting a liquid crystal display image, comprising: calculating an inclination between a display image frame calculated from a value and a scanning line on a screen, and correcting the inclination of an image to be inspected. 6. In a display image displaying an arbitrary pattern, a display image frame or a pattern which is always displayed is scanned along each of m horizontal inspection lines and n vertical inspection lines to obtain a density The position where the sudden change occurs is defined as the boundary position of the display pattern, the tilt value of the display pattern is calculated from the 2 (m + n) boundary position data obtained here, and the tilt of the inspection target image is corrected by the tilt value. Characteristic liquid crystal display image inspection method. 7. In a display image displaying an arbitrary pattern, image data that can always be detected is used as a first inspection position reference, and a second inspection reference is set for each pattern to be inspected, and the second inspection reference is set. A liquid crystal display image inspection method characterized in that an inspection area including a part of an inspection target pattern is defined on the basis of an inspection standard.