• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2002.11a
• /
• pp.529-534
• /
• 2002

출입 관리는 위조 여권 소지자, 수배자, 출입국 금지자 또는 불법 체류자 등의 출입국 부적격자를 검색하고 출입국자를 관리하기 위하여 행하여진다. 한편, 여권에는 사진, 국적, 성명, 주민등록번호, 성별, 여권번호 등을 포함한 정보들로 이루어져 있다. 이러한 출입국 관리 시스템은 출입국 심사 시간이 길어 출입국자에게 불편이 따르고 또한 출입국 부적격자에 대한 정확한 검색이 불분명하여 체계적으로 관리하기가 어렵다. 이러한 종래의 문제점을 개선하기 위해 영상 처리와 문자 인식을 이용한 여권 인증 시스템을 제안한다. 본 논문에서는 여권 영상에 대해 소벨 연산자와 스미어링 기법 그리고 윤곽선 추적 알고리즘을 이용하여 사진영역, 코드 영역 및 개별 코드 문자를 추출하였다. 추출된 개별 코드 인식은 ART2 알고리즘을 기반으로 한 RBF 신경망을 제안하여 여권 인식에 적용하였다. 제안된 방법의 성능을 확인하기 위해서 실제 여권 영상들을 대상으로 실험한 결과, 제안된 방법이 여권 인식에 우수한 성능이 있음을 확인하였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.76-78
• /
• 2022

기계 소비(machine consumption)를 대상으로 하는 이미지 및 비디오의 압축 목표는 사람이 소비하는 것과 다를 수 있다. 예를 들어, 인간의 소비에는 이미지 및 비디오의 전체 캡처 영역이 필요하지만, 기계는 응용 프로그램에서 요구하는 이미지 또는 비디오의 일부만 사용할 수 있다. 이 논문에서는, 효과적인 기계 소비를 위한 영상 압축 방법 개발을 통한 다중 임무(task) 수행 성능 개선 및 압축률(bpp) 향상을 목적으로, 관심 객체(object of interest) 또는 관심 영역(region of interest)을 식별하기 위한 신경망 기반 영상 분석 작업을 수행하는 전처리 네트워크 구성 방법을 제안한다.

• The KIPS Transactions:PartB
• /
• v.11B no.6
• /
• pp.667-674
• /
• 2004

Extraction of texts from images is a fundamental and important problem to understand the images. This paper suggests a text region verification method by statistical means of stroke features of the characters. The method extracts 36 dimensional features from $16\times16$sized text and non-text images using wavelet transform - these 36 dimensional features express stroke and direction of characters - and select 12 sub-features out of 36 dimensional features which yield adequate separation between classes. After selecting the features, SVM trains the selected features. For the verification of the text region, each $16\times16$image block is scanned and classified as text or non-text. Then, the text region is finally decided as text region or non-text region. The proposed method is able to verify text regions which can hardly be distin guished.

• Journal of Advanced Marine Engineering and Technology
• /
• v.39 no.6
• /
• pp.649-655
• /
• 2015

In this paper, we propose a defogging algorithm for a single image. Dark channel prior (DCP), which is a well-known defogging algorithm, can cause halo artifacts on boundary regions, low-contrast defogging images, and requires a large computational time. To solve these problems, we use histogram information with DCP on transmission estimation regions and a multi-resolution method. Local histogram information can reduce the low-contrast problem on a defogging image, and the multi-resolution method with edge information can reduce the total computational time and halo artifacts. We validate the proposed method by performing experiments on fog images, and we confirm that the proposed algorithm is more efficient and superior than conventional algorithms.

• Journal of the Korean Society for information Management
• /
• v.25 no.3
• /
• pp.83-97
• /
• 2008

This thesis presents the methods and tool development of performance assessment as final stage for information literacy instruction of teacher-librarians. The evaluation is an important part in information literacy instruction of teacher-librarians in the school libraries. But until now, there is little research on evaluation methodologies in the instructional activities of teacher-librarians. There is only evaluation domain presented in the primary and secondary 'Information and the Library' curriculum. To develop an evaluation model as a base for measuring performance assessment related to teacher-librarians is important. The goal of this thesis is to provide teacher-librarians with basic principles to develop many kinds of performance assessment tools in their own instructional activities.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.26 no.2
• /
• pp.165-173
• /
• 2008

This study aims at generating automatically three dimensional geometric models of roads using LIDAR data and a digital map. The main processes in the proposed method are (1) generating a polygon encompassing a road region using a road layer from the digital map, (2) extracting LIDAR points within the road region using the polygon, (3) organizing the points into surface patches and grouping the patches into surface clusters, (4) searching the road surface clusters and generating the surface model from the points linked to the clusters, (5) refining the boundary using a digital map. By applying the proposed method to real data, we successfully generated the linear and surface information of the roads.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.1 s.307
• /
• pp.55-66
• /
• 2006

We propose a progressive image retrieval method based on an efficient combination of multiresolution color and torture features in wavelet transform domain. As a color feature, color autocorrelogram of the hue and saturation components is chosen. As texture features, BDIP and BVLC moments of the value component are chosen. For the selected features, we obtain multiresolution feature vectors which are extracted from all decomposition levels in wavelet domain. The multiresolution feature vectors of the color and texture features are efficiently combined by the normalization depending on their dimensions and standard deviation vector, respectively, vector components of the features are efficiently quantized in consideration of their storage space, and computational complexity in similarity computation is reduced by using progressive retrieval strategy. Experimental results show that the proposed method yields average $15\%$ better performance in precision vs. recall and average 0.2 in ANMRR than the methods using color histogram color autocorrelogram SCD, CSD, wavelet moments, EHD, BDIP and BVLC moments, and combination of color histogram and wavelet moments, respectively. Specially, the proposed method shows an excellent performance over the other methods in image DBs contained images of various resolutions.

• Journal of Biomedical Engineering Research
• /
• v.19 no.6
• /
• pp.551-562
• /
• 1998

Amultiresolution filtering algorithm based on the physical SPECT lesion detachability provides and optimal solution for SPECT reconstruction problem. Related to the previous study, we estimated the SPECT lesion detection capability by m minimum detectable lesion sizes (MDLSs), and generated m reconstruction filters which are designed to maximize the smoothing effect at a fixed MDLS-dependent resolution level $\frac{MDLS}{4\sqrt{2In2}}$. The proposed multiresolution filtering algorithm used a coarse-to-fine approach for the m-level resolution filter images obtained from these m filters for a given projection image. First, the local homogeneity is determined for every pixel of the filter images, by comparing the local variance value computed in a window centered at the pixel and the mode determined from the distribution of the local variances. Based on the local homogeneity, the pixels declared as homogeneous are chosen from the filter image of the lowest resolution, and for the other pixels the same process is repeated for the higher resolution filter images. For the non-homogeneous pixels after this pixels after this repetition process ends, the pixel values of the highest resolution filter image are substituted. From the results of the simulated experiments, the proposed multiresolution filtering algorithm showed a strong smoothing effect in the homogeneous regions and a significant resolution improvement near the edge regions of the projection images, and so produced good adaptability effects in the reconstructed images.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.12
• /
• pp.2146-2152
• /
• 2008

In In this paper, we proposed a new JPEG2000 codec using multiresolution-based rendering (MRBR) technique for video compression of 3-dimensional digital cinema. We introduced discrete wavelet transform (DWT) for stereoscopic image and stereo matching technique in the wavelet domain. The disparity was extracted using stereo matching and transmitted with the reference (left) image. Since the generated right image was degraded by the occlusion lesion, the residual image which is generated from difference between the original right image and the generated one was transmitted at the same tine. The disparity data was extracted using the dynamic programming method in the disparity domain. There is high correlation between the higher and lower subbands. Therefore we decreased the calculation amount and enhanced accuracy by restricting the search window and applying the disparity information generated from higher subband.

• Journal of Korea Soil Environment Society
• /
• v.5 no.1
• /
• pp.3-12
• /
• 2000

The multi-region model, to describe preferential flow, is an equation representing solute transport in soils by dividing soil into numerous pore groups and using the hydraulic properties of the soil. As the model partial differential equation (PDE) is solved numerically with finite difference methods. a modified equivalent partial differential equation(MEPDE) of the partial differential equation of the multi-region model is derived to analyze the accuracy and consistency of the solution of the model PDE and the Von Neumann method is used to analyze the stability of the finite difference scheme. The evaluation obtained from the MEPDE indicated that the finite difference scheme was found to be consistent with the model PDE and had the second order accuracy The stability analysis is performed to analyze the model PDE with the amplification ratio and the phase lag using the Von Neumann method. The amplification ratio of the finite difference scheme gave non-dissipative results with various Peclet numbers and yielded the most high values as the Peclet number was one. The phase lag showed that the frequency component of the finite difference scheme lagged the true solution. From the result of the stability analysis for the model PDE, it is analyzed that the model domain should be discretized in the range of Pe < 1.0 and Cr < 2.0 to obtain the more accurate solution.