• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.12
• /
• pp.5990-6008
• /
• 2019

Recognizing license plate from its traffic camera captured images is one of the most important aspects in many traffic management systems. Despite many sophisticated license plate recognition related algorithms available online, license plate recognition is still a hot research issue because license plates in each country all round the world lack of uniform format and their camera captured images are often affected by multiple adverse factors, such as low resolution, poor illumination effects, installation problem etc. A novel method is proposed in this paper to solve the inclination problem of license plates in their camera captured images through four parts: Firstly, special edge pixels of license plate are chosen to represent main information of license plates. Secondly, least square methods are used to compute the inclined angle of license plates. Then, coordinate rotation methods are used to rotate the license plate. At last, bilinear interpolation methods are used to improve the performance of license plate rotation. Several experimental results demonstrated that our proposed method can solve the inclination problem about license plate in visual aspect and can improve the recognition rate when used as the image preprocessing method.

• 보존과학연구
• /
• s.17
• /
• pp.123-160
• /
• 1996

There have been many studies and experiments regarding exhibition lighting. Many experiments on photochemical damaging effect and visibility resulted in a practice limited to assigning light levels and adjusting annual exposure time. The three damaging factors to the artifacts are intrinsicsusceptibility to absorb radiant energy, spectral distribution of light source and intensity of illumination and time of exposure. Dividing all the artfacts into three categories to suggest a recommended illuminance level causes some problems. Blue wool, for example, used as the reference material for susceptibility, is not a standard material representing museum artifacts. In the most light sensitive category, ISO class I or anything below have been excluded. The exposure time of one soure can be three times more than another sourece. The spectral distribution of the light source and the relative spectral responsibility of the artifact are not considered in the practice. So in case of very light sensitive material, the recommended illuminance is only the referring value and it is indispensable to check the characteristic of susceptibility of each artifacts. Daylighting is prevailing method to solve the psychological need of the visitors. However, it sould transparent, and should not diffused, and the green-house effect must be considered. llluminance uniformity should based on the maximum illuminance to handle the limitation of exposure for the conservation of a large sensitive object such as a painting. Damage index is not absolute reference for selecting the lighting source because it is experimented from the paper of low grade then calculated. Visibility should be increased by reducing the visual noiseand by planning of appropriate luminance contrast. This paper reviews the problems with the previous studies and experiment sand the current exhibition lighting design practice. The plan for museum showcase lighting is to check the susceptibility and to raise the visibility simultaneously.

• Journal of Biomedical Engineering Research
• /
• v.27 no.3
• /
• pp.125-130
• /
• 2006

The number of the living endothelial cells and the shape of those are very import clinical parameters for the evaluation of the quality of cornea. In this paper, we developed the automated endothelial cell counting and shape analysis algorithm for a confocal microscope. Since, the endothelial images from the confocal microscope has a non-uniform illumination and low contrast between cell boundaries and cell bodies, it is very difficult to segment the cells from the endothelial images. To cope with these difficulties, we proposed the new two stage image processing algorithm. At first stage algorithm, we used a high-pass filter and histogram equalization to compensate the non-uniform brightness pattern and a morphological filter and a watershed method are applied to detect the boundary of cells. From this stage, we could count the number of cells in an endothelial image. At second stage algorithm, we used a Voronoi diagram method to classify the shape of cells. This cell shape analysis and the percent of hexagonal cells are very sensitive in detecting the early endothelium damage. To evaluate the performance of the proposed system, we p개cessed seven endothelial images obtained using a confocal microscope. The proposed system correctly counted 95.5% cells and classified 92.0% of hexagonal cell shapes. This result is better than any others in this research area.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.5
• /
• pp.437-443
• /
• 2015

This paper presents face recognition based on the fusion of visible image and thermal infrared (IR) texture estimated from the face image in the visible spectrum. The proposed face recognition scheme uses a multi- layer neural network to estimate thermal texture from visible imagery. In the training process, a set of visible and thermal IR image pairs are used to determine the parameters of the neural network to learn a complex mapping from a visible image to its thermal texture in the low-dimensional feature space. The trained neural network estimates the principal components of the thermal texture corresponding to the input visible image. Extensive experiments on face recognition were performed using two popular face recognition algorithms, Eigenfaces and Fisherfaces for NIST/Equinox database for benchmarking. The fusion of visible image and thermal IR texture demonstrated improved face recognition accuracies over conventional face recognition in terms of receiver operating characteristics (ROC) as well as first matching performances.

• Journal of IKEEE
• /
• v.24 no.1
• /
• pp.147-153
• /
• 2020

Conventional lane detection algorithms have problems in that the detection rate is lowered in road environments having a large change in curvature and illumination. The probabilistic Hough transform method has low lane detection rate since it exploits edges and restrictive angles. On the other hand, the method using a sliding window can detect a curved lane as the lane is detected by dividing the image into windows. However, the detection rate of this method is affected by road slopes because it uses affine transformation. In order to detect lanes robustly and avoid obstacles, we propose driving assist system using semantic segmentation based on deep learning. The architecture for segmentation is SegNet based on VGG-16. The semantic image segmentation feature can be used to calculate safety space and predict collisions so that we control a vehicle using adaptive-MPC to avoid objects and keep lanes. Simulation results with CARLA show that the proposed algorithm detects lanes robustly and avoids unknown obstacles in front of vehicle.

• 한국태양에너지학회:학술대회논문집
• /
• 2009.11a
• /
• pp.92-97
• /
• 2009

It is more necessary to consider the various factors for developmenting visible PV module of alternative window than traditional PV module. It must have sufficient performance which is Tvis, daylighting, daylight factor, glare index. so that more needs to consider suitable plan and total evaluated technology. Under the this background. For using commonly a combination BIPV module system and Daylinghting that can alternative architectural window, our goal on this study is drawing proper window area ratio as the window by analyzing lighting performance and glare index depending on transmittance of DSSC. On this study, we drew the result about window area ratio that can apply in the building when applying DSSC in the window. In situation that window is alternated as curtain wall in atrium that has big Widow area, if applying red 15.8% DSSC of low transmittance, it is expect to proper because it is suitable illumination standard and doesn't occur a discomfort glare. In case of office, we propose to apply red 33.2% or blue 35.2% DSSC of high transmittance for no affecting lighting load. we expect to contribute to select proper and effective window when applying the window in the building by drawing the window area ratio that can apply in thee building depending on transmittance of DSSC and offering the glare index data.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.12
• /
• pp.71-82
• /
• 2012

In this paper, we propose a license plate detection method with improved Adaboost learning and MCT (Modified Census Transform). The MCT represents the local structure patterns as integer numbered feature values which has robustness to illumination change and memory efficiency. However, since these integer values are discrete, a lookup table is needed to design a weak classifier for Adaboost learning. Some previous research efforts have focused on minimization of exponential criterion for Adaboost optimization. In this paper, a method that uses MCT and improved Adaboost learning based on Newton's optimization to exponential criterion is proposed for license plate detection. Experimental results on license patch images and field images demonstrate that the proposed method yields higher performance of detection rates with low false positives than the conventional method using the original Adaboost learning.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.3
• /
• pp.679-686
• /
• 2014

Recently, users can get countless images and videos by network. So, the compression technology of image and video is researched more and more. However, the situation which is the interested range of the image is occurred. For instance, since the region of face is more important than background, the image compression technology bases on the region of interest (ROI) is necessary, in the ATM environment. In this research, given the human visual system, which are not sensitive to illumination variations at very dark and light regions of image, we calculate the standard deviation of block and use this value to define the ROI. In encoding process, the relatively high quality can be obtained at the ROI and the relatively low quality can be obtained at the non ROI. In proposed scheme, the feature which is the encoding process according to subjectively image quality can be demonstrated. Finally, this proposed scheme is applied to JPEG standard. The experimental results demonstrate that proposed scheme can achieve better image quality at the high compression ratio.

• Electronic Materials Letters
• /
• v.14 no.6
• /
• pp.669-677
• /
• 2018

In this work, the high ${\kappa}$ $Zr_xAl_{1-x}O_y$ films with a different Zr concentration have been deposited by atomic layer deposition, and the effect of Zr concentrations on the structure, chemical composition, surface morphology and dielectric properties of $Zr_xAl_{1-x}O_y$ films is analyzed by Atomic force microscopy, X-ray diffraction, X-ray photoelectron spectroscopy and capacitance-frequency measurement. The effect of Zr concentrations of $Zr_xAl_{1-x}O_y$ gate insulator on the electrical property and stability under negative bias illumination stress (NBIS) or temperature stress (TS) of ZnSnO (ZTO) TFTs is firstly investigated. Under NBIS and TS, the much better stability of ZTO TFTs with $Zr_xAl_{1-x}O_y$ film as a gate insulator is due to the suppression of oxygen vacancy in ZTO channel layer and the decreased trap states originating from the Zr atom permeation at the $ZTO/Zr_xAl_{1-x}O_y$ interface. It provides a new strategy to fabricate the low consumption and high stability ZTO TFTs for application.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.1
• /
• pp.1-11
• /
• 2021

This paper presents a procedure to extract lane line models using a set of proposed methods. Firstly, an image warping method based on homography is proposed to transform a target image into an image which is efficient to find lane pixels within a certain region in the image. Secondly, a method to use the combination of the results of edge detection and HSL (Hue, Saturation, and Lightness) transform is proposed to detect lane candidate pixels with reliability. Thirdly, erroneous candidate lane pixels are eliminated using a selection area method. Fourthly, a method to fit lane pixels to quadratic polynomials is proposed. In order to test the validity of the proposed procedure, a set of black-box images captured under varying illumination and noise conditions were used. The experimental results show that the proposed procedure could overcome the problems of color-only and edge-only based methods and extract lane pixels and model the lane line geometry effectively within less than 0.6 seconds per frame under a low-cost computing environment.