• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.2
• /
• pp.102-113
• /
• 2014

This paper suggests a novel hardware architecture of a real-time rectification which is to remove vertical parallax of an image occurred in the pre-processing stage of stereo matching. As an off-line step, Matlab Toolbox which was designed by J.Y Bouguet, was used to calculate calibration parameter of the image. Then, based on the Heikkila and Silven's algorithm, rectification hardware was designed. At this point, to enhance the precision of the rectified image, floating-point unit was generated by using Xilinx Core Generator. And, we confirmed that proposed hardware design had higher precision compared to other designs while having the ability to do rectification in real-time.

• Journal of Broadcast Engineering
• /
• v.17 no.5
• /
• pp.851-863
• /
• 2012

Retinex algorithm enhances the contrast of image through visibility improvement. However, the conventional Retinex methods may produces color distortions due to error of hue representation and over-saturation since the methods work in RGB color space. In this paper, we propose a new Retinex algorithm with color correction, which improves contrast by using MSR(Multi-Scale Retinex) working in YCbCr color space and adaptively compensates the color saturation based on the maximum color difference table. Our algorithm maps the color difference signals to the correct gamut to prevent over-saturation phenomenon by considering the correlation between luminance and hue dependent saturation. Simulations results show that the proposed method gives better color improvement compared to the conventional methods.

• The KIPS Transactions:PartB
• /
• v.18B no.2
• /
• pp.57-66
• /
• 2011

To prevent cardiac diseases, quantifying cardiac function is important in routine clinical practice by analyzing blood volume and ejection fraction. These works have been manually performed and hence it requires computational costs and varies depending on the operator. In this paper, an automatic left ventricle segmentation algorithm is presented to segment left ventricle on cardiac magnetic resonance images. After coil sensitivity of MRI images is compensated, a K-mean clustering scheme is applied to segment blood area. A graph searching scheme is employed to correct the segmentation error from coil distortions and noises. Using cardiac MRI images from 38 subjects, the presented algorithm is performed to calculate blood volume and ejection fraction and compared with those of manual contouring by experts and GE MASS software. Based on the results, the presented algorithm achieves the average accuracy of 6.2mL${\pm}$5.6, 2.9mL${\pm}$3.0 and 2.1%${\pm}$1.5 in diastolic phase, systolic phase and ejection fraction, respectively. Moreover, the presented algorithm minimizes user intervention rates which was critical to automatize algorithms in previous researches.

• Journal of Broadcast Engineering
• /
• v.24 no.6
• /
• pp.1152-1162
• /
• 2019

With the rapid development of the VR industry, many VR contents are produced and circulated, and the need for copyright protection is increasing. In this paper, we propose a method of embedding and extracting watermarks in consideration of VR production process. In embedding, SIFT is performed by selecting the region where distortion is minimized in VR production, and transformed into frequency domain using DWT and embedded into the QIM method. In extracting process, in order to correct the distortion in the projection process, the top and bottom regions are changed to different projection methods and some middle regions are rotated using 3DoF to extract the watermark. After this processing, extracted watermark has higher accuracy than the conventional watermark method, and the validity of the proposed watermark is shown by showing that the accuracy is maintained even in various attacks.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.6
• /
• pp.27-34
• /
• 2011

In this paper, we propose an active 3D shape acquisition method based on photometric stereo using camera and flash on a smartphone. Two smartphones are used as the master and slave, in which the slave projects illumination from different locations while the master captures the images and processes photometric stereo algorithm to reconstruct 3D shape. In order to reduce the error, the smartphone's camera is calibrated to overcome the effect of the lens distortion and nonlinear camera sensor response. We apply 5-point algorithm to estimate the pose between smartphone cameras and then estimate lighting direction vector to run the photometric stereo algorithm. Experimental result shows that the proposed system enables us to use smartphone as a 3D camera with low cost and high quality.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.19 no.10
• /
• pp.1175-1184
• /
• 2008

Delay-sum back projection(DSBP) algorithm and the time reversal algorithm based on the finite-difference time-domain method are compared. The two algorithms, which operate in the time domain, can process the ultra-wideband (UWB) radar data to generate images that are close to the original location and shape of the target. For the experiment, the UWB radar consists of a network analyzer, a resistive V dipole antenna, a scanner, and a control computer. The radar aperture is synthesized by linearly scanning the antenna. A calibration procedure is applied to the measured data to remove signal distortion and clutter. The two algorithms are applied to the same data on the same platform. It is shown that the DSBP algorithm produces better images but takes longer time to produce the images than the FDTD-TR algorithm.

• The Journal of Korean Society for Radiation Therapy
• /
• v.29 no.1
• /
• pp.49-56
• /
• 2017

Purpose: To evaluate the accuracy of the Iterative Metal Artifact Reduction (IMAR) algorithm in correcting CT (computed tomography) images distorted due to a metal artifact and to evaluate the usefulness when proton therapy plan was plan using the images on which IMAR algorithm was applied. Materials and Methods: We used a CT simulator to capture the images when metal was not inserted in the CIRS model 062 Phantom and when metal was inserted in it and Artifact occurred. We compared the differences in the CT numbers from the images without metal, with a metal artifact, and with IMAR algorithm by setting ROI 1 and ROI 2 at the same position in the phantom. In addition, CT numbers of the tissue equivalents located near the metal were compared. For the evaluation of Rando Phantom, CT was taken by inserting a titanium rod into the spinal region of the Rando phantom modelling a patient who underwent spinal implant surgery. In addition, the same proton therapy plan was established for each image, and the differences in Range at three sites were compared. Results: In the evaluation of CIRS Phantom, the CT numbers were -6.5 HU at ROI 1 and -10.5 HU at ROI 2 in the absence of metal. In the presence of metal, Fe, Ti, and W were -148.1, -45.1 and -151.7 HU at ROI 1, respectively, and when the IMAR algorithm was applied, it increased to -0.9, -2.0, -1.9 HU. In the presence of metal, they were 171.8, 63.9 and 177.0 HU at ROI 2 and after the application of IMAR algorithm they decreased to 10.0 6,7 and 8.1 HU. The CT numbers of the tissue equivalents were corrected close to the original CT numbers except those in the lung located farthest. In the evaluation of the Rando Phantom, the mean CT numbers were 9.9, -202.8, and 35.1 HU at ROI 1, and 9.0, 107.1, and 29 HU at ROI 2 in the absence, presence of metal, and in the application of IMAR algorithm. The difference between the absence of metal and the range of proton beam in the therapy was reduced on the average by 0.26 cm at point 1, 0.20 cm at point 2, and 0.12 cm at point 3 when the IMAR algorithm was applied. Conclusion: By applying the IMAR algorithm, the CT numbers were corrected close to the original ones obtained in the absence of metal. In the beam profile of the proton therapy, the difference in Range after applying the IMAR algorithm was reduced by 0.01 to 3.6 mm. There were slight differences as compared to the images absence of metal but it was thought that the application of the IMAR algorithm could result in less error compared with the conventional therapy.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2015.07a
• /
• pp.464-465
• /
• 2015

본 논문은 HEVC(High Efficiency Video Coding) 기반의 3차원 비디오 부호기에서 깊이 비디오 부호화의 효율 증대를 위한 디블록킹 필터(deblocking filter)를 제안한다. 디블록킹 필터는 블록 왜곡(blocking artifact)을 보정하기 위한 필터인데 원래 색상 영상의 특성에 맞게 설계되어서 비슷한 목적을 지닌 SAO(Sample Adaptive Offset)와 더불어 기존 방법의 깊이 비디오 부호화에서는 사용되지 않는다. 제안 방법은 디블록킹 필터의 사전 실험 통계에 기반하여 기여도가 낮은 normal 필터를 제외시킨다. 또한, 깊이 비디오의 특성을 고려하여 임펄스 응답(impulse response)를 변형하였다. 이 변형된 디블록킹 필터를 깊이 비디오 부호화에만 적용하고 색상 비디오 부호화에는 기존 디블록킹 필터를 사용하였다. 3D-HTM(HEVC Test Model) 13.0 참조 소프트웨어에 구현하여 실험한 결과, 기존 방법에 비해 깊이 비디오 부호화 성능이 5.2% 향상되었다. 색상-깊이 비디오 간 참조가 있기 때문에 변형된 깊이 비디오 부호화가 색상 비디오 부호화 효율에 영향을 끼칠 수도 있지만 실험 결과 색상 비디오 부호화 성능은 유지되었다. 따라서 제안 방법은 성공적으로 깊이 비디오 부호화의 효율을 증대시켰다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2012.05a
• /
• pp.28-29
• /
• 2012

IT 및 자동차 관련 기술의 융합 기술의 발전에 따라 자동차의 안전 및 운전편의 정보 제공에 대한 관심이 높아지면서 인간의 시각 및 지각의 한계를 보완해 줄 수 있는 보조 도구들에 대한 연구 및 개발이 활발히 이루어지고 있다. 그러나 기존의 컴퓨터 비전 기반 도로 교통 표지판 자동 검출 및 인식 기술들은 센싱 장비들의 가격 또는 조도와 원근감에 따른 교통 표지판의 색상과 모양 왜곡으로 인해 해당 표지판들의 검출을 어렵게 한다. 따라서 본 논문에서는 차량 내 탑재가 가능한 저가의 비전 카메라를 이용하여 교통표지판 칼라 분석 및 원근 보정을 통해 운전자에게 효과적으로 도로표지판 정보를 제공할 수 있는 검출기법을 제안한다. 실험을 통해 도로주행영상 내 도로표지판들을 효과적으로 검출할 수 있음을 확인하였다.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_1
• /
• pp.1219-1230
• /
• 2022

Multi-spectral drones in agricultural observation require quantitative and reliable data based on physical quantities such as radiance or reflectance in crop yield analysis. In the case of remote sensing data for crop monitoring, images taken in the same area over time-series are required. In particular, biophysical data such as leaf area index or chlorophyll are analyzed through time-series data under the same reference, it can be directly analyzed. So, comparable reflectance data are required. Orthoimagery using drone images, the entire image pixel values are distorted or there is a difference in pixel values at the junction boundary, which limits accurate physical quantity estimation. In this study, reflectance and vegetation index based on drone images were calculated according to the correction method of drone images for time-series crop monitoring. comparing the drone reflectance and ground measured data for spectral characteristics analysis.