KR100570268B1 - method for detecting irregular mura in a light-related plate element for a flat panel using a feature strength - Google Patents

Abstract

According to the present invention, the brightness distribution of the entire image is not constant in the flat panel light-related plate element by using the feature point intensity in the image obtained from the light-related plate element for flat panel, and the luminance distribution of the cross section has a nonlinear structure of more than a cubic curve. The present invention provides a method for efficiently and automatically detecting irregularly shaped stains having the most frequent occurrence among stain defects.

The irregularity spot detection method is a defect detection method for detecting irregularity spots in a defective state through image processing by acquiring an image from a light-related plate element for a flat panel, wherein 90% of all pixels are obtained from a histogram of the acquired image. Performing normalization on brightness based on the existing level value; Applying the normalized image to the averaging filter as a whole to reduce the effect of noise; Extracting feature points while performing line-by-line scanning in consideration of the overall pattern of the image; Investigating the degree of clustering and removing feature points erroneously detected by noise in the normal image; And performing normalization to obtain an indefinite parameter value using all remaining feature points and describing it linearly, and then quantizing the parameter value that should be a final criterion to appear at a linear level. It is characterized by.

Description

Method for detecting irregular mura in a light-related plate element for a flat panel using a feature strength}

1 is a flow chart showing a method for detecting irregularities of the light-related plate element for a flat panel using the feature point intensity in accordance with an embodiment of the present invention.

<Description of the symbols for the main parts of the drawings>

S1: video input step S2: histogram adjustment step

S3: Averaging filtering step S4: Feature point extraction step

S5: Computation of clusters of feature points S6: Quantization of clusters

S7: Level measurement step S8: Result output storage step

The present invention relates to a method for detecting irregularity unevenness of a flat panel light-related plate element, and more particularly, in a flat panel light-related plate element by using the characteristic point intensity in an image obtained from the flat panel light-related plate element. The luminance distribution of the whole image is not constant, and the luminance distribution of the cross section itself has a nonlinear structure of more than a cubic curve, and it is possible to efficiently and automatically detect irregularities that have the highest frequency of unevenness. .

Currently, flat panels, such as LCDs and PDPs, are constructed by stacking / incorporating light-related plate elements such as glass substrates or optical pattern plates. appear. In other words, various defects such as uneven color, dust, cracks, scratches, and other defects such as point defects, predecessors, and certain types of linear stains such as block shapes, and sublinear stains with undefined edges may appear. do.

In order to discriminate the defects of the optically related plate element, inspection and discrimination are usually carried out by visual inspection, which not only lowers the productivity and accuracy of the inspection, but also consumes a lot of cost and lacks reproducibility in detecting the defects. Therefore, in order to obtain objective and cost-effective effects, it is necessary to introduce an automatic inspection system.

In this way, the automatic inspection system is capable of reproducible processing in real time using advanced computer vision technology and high-speed image processing algorithms, thereby improving productivity as well as efficient quality control. However, there are some difficulties in using image processing by acquiring an image at a high resolution. First, the luminance level of the panel itself appears nonlinearly over the entire surface, and the luminance according to the activation time even after a signal is applied to the panel. The distribution will change. Second, each model has different image characteristics according to the position of the back light and the internal structure of the diffusion sheet. Due to these difficulties, it is difficult to apply image segmentation techniques such as general edge operators.

 Accordingly, the present invention is to solve this problem, and relates to an image processing algorithm for automatic inspection of semi-finished or finished products for flat panel displays (FPD), a high level of inspection process that relies on the existing visual inspection Its main purpose is to enable efficient quality control as well as productivity by reproducible processing in real time using computer vision technology and high-speed image processing algorithms.

To this end, the present invention, the brightness distribution of the entire image is not constant in the flat panel light-related plate element by using the feature point intensity in the image obtained from the light-related plate element for flat panel, the luminance distribution of the cross-section itself is more than the cubic curve The present invention provides a method for efficiently and automatically detecting irregularly shaped stains having a non-linear structure and having the highest frequency of stain defects.

In order to achieve the above object, the method of detecting irregularities in the flat panel optically related plate element using the feature point intensity according to an embodiment of the present invention is obtained through image processing by obtaining an image from the flat panel optically related plate element. In the defect detection method of the light-related plate element for flat panel for detecting irregular formation of the medium, the step of normalizing the brightness based on the level value of 90% or more of all pixels from the histogram of the obtained image ; Applying the normalized image to the averaging filter as a whole to reduce the effect of noise; Extracting feature points while performing line-by-line scanning in consideration of the overall pattern of the image; Investigating the degree of clustering and removing feature points erroneously detected by noise in the normal image; And performing normalization to obtain an indefinite parameter value using all remaining feature points and describing it linearly, and then quantizing the parameter value that should be a final criterion to appear at a linear level. It is characterized by.

Flat panel light-related plate elements that can be inspected include TFT-LCD finished products, LCD semi-finished products such as BLU and diffuser plates, organic EL, and PDP, and can also be applied to ITO glass and optical film series. The image related to the flat panel product is not only difficult to acquire, but the acquired image is high resolution and has non-linear luminance change in the image. Therefore, in the present invention, the use of pre-processing and post-processing techniques for image improvement, and high-speed inspection algorithm for real-time inspection It is characterized by.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a flowchart illustrating a method for detecting irregularity spots of an optically related plate element for a flat panel according to an embodiment of the present invention.

In FIG. 1, the method for detecting irregularities of the flat panel optically related plate element of the present invention basically includes a histogram adjusting step (step S2), an averaging filtering step (step S3), a feature point extraction step (step S4), and a feature point cluster. And a cluster quantization step (step S6).

In an embodiment of the present invention illustrated in FIG. 1, since the average luminance is changed according to a model such as LCM as the image input in step S1 as a preprocessing level, at least 90% of all pixels exist from the histogram of the acquired image. Normalize brightness by value. And even in the normal image, the center of the acquired image is relatively bright compared to the whole, so it is sensitive to the effects of camera noise, etc., which can be detected later when extracting feature points, which can cause false positive errors. The image is applied to the averaging filter as a whole to reduce the influence of noise (steps S2 and S3).

Thereafter, the acquired image is generally bright in the center of the image and becomes darker toward the edges, and maintains a lower pattern at the lower portion than the upper portion. The feature points are extracted while scanning each line in consideration of the overall pattern of the image (step S4). ). Although the averaging filter was previously applied, the feature points due to noise may still exist in the normal image, and the irregularity spots have some clustering characteristics. It is excluded (step S5). Typically, the cluster figure identifies the number of feature points located in a specific block based on the feature points, and excludes the final result from the case where the number does not meet the appropriate criteria.

Since the reliability of the feature points extracted by the above-described process has been increased to some extent, the negated parameters are obtained using all remaining feature points, and then quantized (step S6). The degree of irregularity can be described as the mean of the entire image, the level difference around the feature points detected in the variance and irregularity speckles, and finally the cluster in the two-dimensional block surrounding the feature points. The parameter is calculated using a formula using all these values, and the obtained parameter values do not represent the degree of negation linearly, so normalization is performed to describe them linearly. The parameter value, which should be the final criterion, should appear at a linear level, and the calculated parameter value is appropriately quantized.

In addition, the quantized parameter is a given image reference, and its level is measured (step S7), and can be newly quantized and expressed when the brightness / size of the image changes. Then, the results regarding such irregularity staining are output or stored in step S8.

On the other hand, the method of the present invention can be carried out by systems which are similar to the inspection apparatus of the light-related plate element for the sensitive flat panel developed by the applicant and filed in application No. 2002-11358 dated March 4, 2002. have.

According to the configuration and operation of the method for detecting irregularities in the flat panel light-related plate element using the feature point intensity according to the embodiment of the present invention, the feature point intensity is used in the image obtained from the flat panel light-related plate element. Therefore, the luminance distribution of the whole image is not constant in the flat panel light-related plate elements, and the luminance distribution of the cross section itself has a nonlinear structure of more than a cubic curve. It is possible to detect such effects.

Claims (3)

delete delete Normalization of brightness based on the level value of 90% or more of the total pixels from the histogram of the acquired image, in order to detect an image irregularity unevenness through image processing by acquiring an image from the light-related plate element for flat panel Performing the step of applying the image, applying the image normalized to reduce the influence of the noise to the averaging filter as a whole, extracting the feature points while scanning each line in consideration of the overall pattern of the image, and examining the clustering degree. After removing the feature points detected by noise in the normal image, normalizing to obtain the negated parameter value using all remaining feature points and describing them linearly, the parameter to be the final criterion A flat constructed comprising quantizing the values to appear at a linear level Neolyong in negative form stain detection method for an optical element associated plate, The step of removing the feature point is to grasp the number of feature points located in a specific block based on the feature point as a cluster and excludes the result from the final result when the number does not meet the appropriate criteria. Of the light-related plate element for a flat panel using a feature point intensity, characterized by a mean, a level difference around a feature point detected in scattering and irregularity spots, and finally a degree of irregularity described as a cluster in a two-dimensional block surrounding the feature point. Method for detecting irregularity stains.

Images