Il Choi

Inspection of defect on LCD panel using polynomial approximation

Nowadays, the CRT display devices are rapidly substituted for LCD display devices. In LCD panel production, the inspection process relies on human vision. So, the results of inspection is likely to be subjective. Therefore, we need the automated inspection system for LCD devices by obtaining objective decision level. In this paper, we propose an efficient automated inspection algorithm using polynomial approximation and optimal thresholding technique. The experimental result shows us that our proposed system can have enough performance for substitution of human inspectors.

Face and Facial Landmarks Location Based on Log-Polar Mapping

In this paper, a new approach using a single log-polar face template has been adopted to locate a face and its landmarks in a face image with varying scale and rotation parameters. The log-polar mapping which simulates space-variant human visual system converts scale change and rotation of input image into constant horizontal and cyclic vertical shift in the log-polar plane. The intelligent use of this property allows the proposed method to eliminate the need of using multiple templates to cover scale and rotation variations of faces and thus achieves efficient and reliable location of faces and facial landmarks.

A generalized symmetry transform with selective attention capability for specific corner angles

A generalized symmetry transform evaluates evidence of local symmetry, yet it has no selective attention capability for corner angles. Therefore, a symmetry transform is proposed that can recognize symmetric features tuned to specific corner angles. Experimental results with synthetic and real imagery confirm the effectiveness of the proposed transform in discriminating polygonal objects.

Stereo System for Tracking Moving Object Using Log-Polar Transformation and Zero Disparity Filtering

An active stereo-vision system enables a target object to be localized based on passing small disparities without heavy computation to identify the target. However, this simple method is not applicable to situations where a distracting background is included or the target and other objects are simultaneously located in the zero disparity area. Accordingly, to alleviate these problems, the current study combined filtering and foveation, which employs high resolution in the center of the visual field, while suppressing the periphery. An image pyramid and log-polar transformation are compared for the foveated image representation. The stereo disparity of the target is also extracted using projection to maintain a small stereo disparity during tracking. Experiments demonstrated that a log-polar transformation was superior to both the image pyramid and the traditional method for separating a target from a distracting background, and comparatively enhanced the tracking performance.

Motion-Based Boundary Tracking of Moving Object Using Parametric Active Contour Model

This paper presents a motion-based boundary tracking method for a moving deformable object in an image sequence using a parametric active contour model. Deciding the local converging directions of the contour points is essential for correctly extracting the boundary of a moving deformable object. Thus, a new energy function for a parametric active contour model is proposed based on the addition of a directional energy term using a frame difference map to the greedy snake. The frame difference map is used to obtain motion information on an object with fast and non-rigid motion. Plus, updating rules for the frame difference map are also developed to encourage the stable convergence of the contour points. Experiments on a set of synthetic and real image sequences show that the proposed method could fully track a speedy deformable object while exactly extracting the boundary of the object in every frame.

IR Image Segmentation using GrabCut

This paper proposes a method for segmenting objects from the background in IR(Infrared) images based on GrabCut algorithm. The GrabCut algorithm needs the window encompassing the interesting known object. This procedure is processed by user. However, to apply it for object recognition problems in image sequences. the location of window should be determined automatically. For this, we adopted the Otsu` algorithm for segmenting the interesting but unknown objects in an image coarsely. After applying the Otsu` algorithm, the window is located automatically by blob analysis. The GrabCut algorithm needs the probability distributions of both the candidate object region and the background region surrounding closely the object for estimating the Gaussian mixture models(GMMs) of the object and the background. The probability distribution of the background is computed from the background window, which has the same number of pixels within the candidate object region. Experiments for various IR images show that the proposed method is proper to segment out the interesting object in IR image sequences. To evaluate performance of proposed segmentation method, we compare other segmentation methods.

Design of Double Layer Shielding Structure using Eco-friendly Shielding Materials

Recently, a barium(Ba) and an iodine(I) being studied as a conventional lead(Pb) alternatives of shielding material has excellent shieding rate, but the characteristic x-ray photons in the energy range near 30 keV line is released. In this study, with bismuth oxide(BiO) coupled barium sulfate(BaSO4) double layer, transmitted spectra, shieding rates and relative weighting rates were evaluated to validate the applicability of eco-friendly double layer shieding structure using monte carlo simulations as a prior study. From the evaluation results, in 0.4mm and 0.5mm thickness of BiO layers coupled with top 0.1mm-BaSO4 layer, the shieding rate showed 1.9% and 3.9% higher than 0.6mm thickness of Pb single layer, respectively. In addition, the relative weight also 28% and 34.5% lower than 0.6mm-Pb in 0.4mm and 0.5mm thickness of BiO layers coupled with top 0.1mm-BaSO4 layer.

The Detection of Curve-type Defects in the TFT-LCD panels with Machine Vision

The needs of automated inspection system are grown for the flat panel displays (FPD). However, it is so difficult to detect the defects on the surface of FPD due to non-uniform intensity distribution through the whole panel. Especially, curve-type defects such as micro sized thread and scratch are too difficult to detect with the conventional methods. In this paper, we propose a robust detection algorithm for curve-type defects. The proposed algorithm is composed of flattening, segmentation, and classification. The flattening makes intensity distribution likely to be uniform. The segmentation extracts candidates of curve-type defect and the classification selects real defects from candidates. The experimental result verifies the performance of the proposed method.