• 한국멀티미디어학회논문지
• /
• 제10권10호
• /
• pp.1271-1283
• /
• 2007

평판 디스플레이(FPD)의 결함 검출은 패널 영상의 불균일한 휘도 변화로 인해 정확한 결함 검출이 어렵다. 본 논문은 FPD 패널 영상의 휘도변화를 B-스플라인 표면으로 근사화하고, 이로부터 다양한 결함을 검출하는 방법을 제안한다. B-스플라인 표면 근사화시 잡음 및 결함에 해당하는 고주파 부분을 제외하고, 불균일 휘도 변화에 해당하는 배경부분만으로 구성하기 위해, 웨이브릿 변환 후 저주파대역만을 이용한다. 이는 B-스프라인 표면 근사화의 단점인 시간 소모를 획기적으로 줄일 뿐 아니라, 정확성을 향상시키는 결과를 가져온다. 최소의 부대역에서 근사화된 영상은 웨이브릿 합성 과정을 거쳐 원영상의 크기로 재구성되고, 원 영상에서 이를 뺀 차영상이 바로 불균일 휘도의 배경을 보상한 평평한 영상이 된다. 따라서 결과 영상에 단순 문턱치를 이용하여 결함 영역을 쉽게 검출할 수 있으며, 거짓 결함을 제거하기 위해 블랍 해석이 후처리로서 수행된다 또한 인라인 시스템에 적용하기 위해 웨이브릿 변환을 리프팅 기반의 알고리즘으로 구현하여 필름 같은 대용량의 데이터를 고속으로 처리할 수 있게 함으로써, 처리 시간을 크게 감소 시켰다.

• International Journal of Control, Automation, and Systems
• /
• 제6권5호
• /
• pp.746-754
• /
• 2008

This paper proposes a method to detect calibration patterns for accurate camera calibration under complicated backgrounds and uneven lighting conditions of industrial fields. Required to measure object dimensions, the preprocessing of camera calibration must be able to extract calibration points from a calibration pattern. However, industrial fields for visual inspection rarely provide the proper lighting conditions for camera calibration of a measurement system. In this paper, a probabilistic criterion is proposed to detect a local set of calibration points, which would guide the extraction of other calibration points in a cluttered background under irregular lighting conditions. If only a local part of the calibration pattern can be seen, input data can be extracted for camera calibration. In an experiment using real images, we verified that the method can be applied to camera calibration for poor quality images obtained under uneven illumination and cluttered background.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제10권1호
• /
• pp.47-60
• /
• 2006

평판 디스플레이 장치(FPD)의 패널 표면 결함 검출에서 일반적으로 사용되는 단순 문턱값에 의한 결함 검출은 FPD 패널 영상의 불균일한 휘도 변화로 인하여 정확한 결함 검출이 어렵다. 본 논문에서는 이러한 불균일한 휘도 변화를 보상하고, 정확한 결함 검출을 위해 다 해상도 분석방법인 웨이블릿 변환에 기반하여 높은 고주파 잡음제거와 함께 낮은 저주파를 제거함으로써 불균일한 휘도 변화를 보상할 수 있는 알고리즘을 제안하고 구현 하였다. 특히 제조 공정에서의 결함 검출을 실시간 인라인으로 적용하기 위해 리프팅 기반 고속 알고리즘으로 구현하였다.

• Journal of Information Processing Systems
• /
• 제18권4호
• /
• pp.575-586
• /
• 2022

Uneven light in real-world causes visual degradation for underexposed regions. For these regions, insufficient consideration during enhancement procedure will result in over-/under-exposure, loss of details and color distortion. Confronting such challenges, an unsupervised low-light image enhancement network is proposed in this paper based on the guidance of the unpaired low-/normal-light images. The key components in our network include super-resolution module (SRM), a GAN-based low-light image enhancement network (LLIEN), and denoising-scaling module (DSM). The SRM improves the resolution of the low-light input images before illumination enhancement. Such design philosophy improves the effectiveness of texture details preservation by operating in high-resolution space. Subsequently, local lightness attention module in LLIEN effectively distinguishes unevenly illuminated areas and puts emphasis on low-light areas, ensuring the spatial consistency of illumination for locally underexposed images. Then, multiple discriminators, i.e., global discriminator, local region discriminator, and color discriminator performs assessment from different perspectives to avoid over-/under-exposure and color distortion, which guides the network to generate images that in line with human aesthetic perception. Finally, the DSM performs noise removal and obtains high-quality enhanced images. Both qualitative and quantitative experiments demonstrate that our approach achieves favorable results, which indicates its superior capacity on illumination and texture details restoration.

• 한국콘텐츠학회논문지
• /
• 제9권8호
• /
• pp.9-17
• /
• 2009

자동차 번호 인식에 대한 연구가 활발히 이루어져 왔으나, 번호판 이진화 방법들에 대한 비교 연구는 거의 이루어지지 않았다. 이로 인하여 관련 연구자들마다 효과적인 이진화 방법을 찾기 위하여 유사한 시행착오를 겪어 왔다. 본 연구에서는 이러한 시행착오를 줄일 수 있도록 기존의 번호판 이진화 방법들을 구현하여 성능을 양적으로 비교 제시하였다. 이진화 성능 측정은 저수준 척도와 고수준 척도를 모두 사용함으로써 이진화 자체에 대한 평가뿐만 아니라 후속 단계에서의 유용성을 함께 고려하였다. 그리고 이진화 방법들의 특성을 파악하기 위하여 조도의 특성에 따라 번호판 영상을 세 그룹으로 분류하여 이진화 성능을 측정하였다. 실험 결과 조도의 강도보다는 조도의 균일성 여부가 이진화 성능에 더 큰 영향을 미치는 것으로 나타났다. 조도가 균열한 영상은 Otsu의 방법이 가장 효과적이었으며, 조도가 불균일한 영상은 파라미터를 보정한 Niblack 방법이 가장 좋은 결과를 나타냈다.

• ETRI Journal
• /
• 제35권3호
• /
• pp.491-500
• /
• 2013

A connected component analysis from a binary image is a popular character segmentation method but occasionally fails to segment the characters owing to image noise and uneven illumination. A multimethod binarization scheme that incorporates two or more binary images is a novel solution, but selection of binarization methods has never been analyzed before. This paper reveals the best combination of binarization methods and parameters and presents an in-depth analysis of the multimethod binarization scheme for better character segmentation. We carry out an extensive quantitative evaluation, which shows a significant improvement over conventional single-method binarization methods. Experiment results of six binarization methods and their combinations with different test images are presented.

• 전기학회논문지
• /
• 제57권2호
• /
• pp.307-313
• /
• 2008

In this paper, we propose a novel defects inspection algorithm for TFT-LCD panels. We first compensate the distorted image caused by the camera distortion and the uneven illumination environment using the least squares method and the bezier surface. We find a starting point of each pattern for restricting each pattern. A clean image is compared to each pattern to find defects using modified PCSR-G algorithm. The simulation example shows that our algorithm not only inspects the defects well, but also is robust to the 1-pixel error.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2007년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.149-150
• /
• 2007

In this paper, we propose a novel defects inspection algorithm for TFT-LCD panels. We first compensate the distorted image caused by the camera distortion and the uneven illumination environment using the least squares method and the bezier surface. We find a starting point of each pattern. The reference frame, made by subtract method using several clean patterns, is compared to each pattern to find defects. The simulation example shows that our algorithm not only inspects the defects well, but also is robust to the 1-pixel error.

• 스마트미디어저널
• /
• 제2권1호
• /
• pp.31-36
• /
• 2013

A robust and efficient text extraction is very important for an accuracy of Optical Character Recognition (OCR) systems. Natural scene images with degradations such as uneven illumination, perspective distortion, complex background and multi color text give many challenges to computer vision task, especially in text extraction. In this paper, we propose a method for extraction of the text in signboard images based on a combination of mean shift algorithm and weighting scheme of hue and saturation in HSI color space for clustering algorithm. The number of clusters is determined automatically by mean shift-based density estimation, in which local clusters are estimated by repeatedly searching for higher density points in feature vector space. Weighting scheme of hue and saturation is used for formulation a new distance measure in cylindrical coordinate for text extraction. The obtained experimental results through various natural scene images are presented to demonstrate the effectiveness of our approach.

• 한국정보처리학회논문지
• /
• 제6권11호
• /
• pp.3165-3177
• /
• 1999

With the recent outstanding advance in computer software and hardware, a number of researches to enhance the manufacturing speed and the process accuracy has been undertaken in many fields of textile industry. Frequently issued problems of automatic recognition of textile wrinkles in a grey scale image are as follows. First, changes in grey level intensity of wrinkles are so minute. Second, as both colors and patterns in a grey scale image appear in grey level intensity, it is difficult to sort out the wrinkle information only. Third, it is also difficult to distinguish grey level intensity changed by wrinkles from those by uneven illumination. This paper suggests a method of automatic recognition of textile wrinkles that can solve above problems concerned with wrinkles, which can be raised in a manufacturing process as one of errors. In this paper, we first make the outline of wrinkles distinctly, apply the block matching algorithm used in motion estimation, and then estimate block locations of target images corresponding to blocks of standard images with the assumption that wrinkles are kind of textile distortions caused by directional forces. We plot a "wrinkle map" considering distances between wrinkles as depths of wrinkles. But because mismatch can occur by different illumination intensity and changes in tensions and directions of the force, there are also undesirable patterns in the map. Post processing is needed to filter them out and get wrinkles information only. We use average grey level intensity of wrinkle map to recognize wrinkles. When it comes to textile with colors and patterns, previous researches on wrinkles in grey scale image hasn't been successful. But we make it possible by considering wrinkles as distortion.istortion.