• 한국멀티미디어학회논문지
• /
• 제20권7호
• /
• pp.994-1003
• /
• 2017

Ground-glass opacity nodules(GGNs) in chest CT images are associated with lung cancer, and have a different malignant rate depending on existence of solid component in the nodules. In this paper, we propose a method to classify pure GGNs and part-solid GGNs using multiview images and texture analysis in pulmonary GGNs with solid components of 5mm or smaller. We extracted 1521 features from the GGNs segmented from the chest CT images and classified the GGNs using a SVM classification model with selected features that classify pure GGNs and part-solid GGNs through a feature selection method. Our method showed 85% accuracy using the SVM classifier with the top 10 features selected in the multiview images.

• 대한핵의학회지
• /
• 제23권1호
• /
• pp.19-25
• /
• 1989

To assess the usefulness of myocardial SPECT imaging to detect post-myocardial infarction ventricular aneurysms, we analyzed the Technetium-99m MIBI myocardial SPECT images of 16 patients with anterior and/or apical infarction, 9 had the previously reported findings of failure of convergence of the left ventricular walls toward the apex on SPECT images and 8 of them also had ventricular aneurysms. The ventriculography of the 2 patients with mixed pattern revealed 1 case of ventricular aneurysm and 1 case without aneurysm. Among the remaining 5 pateints with converging pattern, none had ventricular aneurysm. Of the other 11 pateints with inferior and/or lateral wall infarction, 1 patient had ventricular aneurysm and the SPECT image couldn't detect the aneurysm. $Department of Internal Medicine, College of Medicine, Seoul National University$ myocardial SPECT images for the detection of ventricular aneurysm had a sensitivity of 90 %, a specificity of 88%, and an accuracy of 89%. Thus we could get the information about presence of ventricular aneurysm as well as the status of the myocardial perfusion from the Tc-99m MIBI myocardial SPECT images.

• Imaging Science in Dentistry
• /
• 제42권1호
• /
• pp.25-33
• /
• 2012

Purpose : This study was performed to determine the accuracy of linear measurements on three-dimensional (3D) images using multi-detector computed tomography (MDCT) and cone-beam computed tomography (CBCT). Materials and Methods : MDCT and CBCT were performed using 24 dry skulls. Twenty-one measurements were taken on the dry skulls using digital caliper. Both types of CT data were imported into OnDemand software and identification of landmarks on the 3D surface rendering images and calculation of linear measurements were performed. Reproducibility of the measurements was assessed using repeated measures ANOVA and ICC, and the measurements were statistically compared using a Student t-test. Results : All assessments under the direct measurement and image-based measurements on the 3D CT surface rendering images using MDCT and CBCT showed no statistically difference under the ICC examination. The measurements showed no differences between the direct measurements of dry skull and the image-based measurements on the 3D CT surface rendering images (P>.05). Conclusion : Three-dimensional reconstructed surface rendering images using MDCT and CBCT would be appropriate for 3D measurements.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제27권1호
• /
• pp.56-74
• /
• 2023

In order to obtain practical and high-quality satellite images containing high-frequency components, a large aperture optical system is required, which has a limitation in that it greatly increases the payload weight. As an attempt to overcome the problem, many multi-aperture optical systems have been proposed, but in many cases, these optical systems do not include high-frequency components in all directions, and making such an high-quality image is an ill-posed problem. In this paper, we use deep learning to overcome the limitation. A deep learning model receives low-quality images as input, estimates the Point Spread Function, PSF, and combines them to output a single high-quality image. We model images obtained from three rectangular apertures arranged in a regular polygon shape. We also propose the Modulation Transfer Function Loss, MTF Loss, which can capture the high-frequency components of the images. We present qualitative and quantitative results obtained through experiments.

• Korean Journal of Geomatics
• /
• 제5권1호
• /
• pp.35-42
• /
• 2005

Precise geometric registration is required in multi-source data fusion process to obtain synergistic results successfully. However, most of the previous studies focus on the assumption of perfect registration or registration in a limited local area with intuitively derived simple geometric model. In this study, therefore, we developed a robust method for geometric registration based on a systematic model that is derived from the geometry associated with the data acquisition processes. The key concept of the proposed approach is to utilize smooth planar patches extracted from LIDAR data as control surfaces to adjust exterior orientation parameters of the aerial images. Registration of the simulated LIDAR data and aerial images was performed. The experimental results show that the RMS value of the geometric discrepancies between two data sets is decreased to less than ${\pm}0.30\;m$ after applying suggested registration method.

• 대한의용생체공학회:의공학회지
• /
• 제26권3호
• /
• pp.133-138
• /
• 2005

This paper presents a new method for measuring ocular torsion using the optical flow. Images of the iris were cropped and transformed into rectangular images that were orientation invariant. Feature points of the iris region were selected from a reference and a target image, and the shift of each feature was calculated using the iterative Lucas-Kanade method. The feature points were selected according to the strength of the corners on the iris image. The accuracy of the algorithm was tested using printed eye images. In these images, torsion was measured with $0.15^{\circ}$ precision. The proposed method shows robustness even with the gaze directional changes and pupillary reflex environment of real-time processing.

• Journal of Radiation Protection and Research
• /
• 제44권4호
• /
• pp.149-155
• /
• 2019

Background: Magnetic resonance (MR) image guided radiation therapy system, enables real time MR guided radiotherapy (RT) without additional radiation exposure to patients during treatment. However, MR image lacks electron density information required for dose calculation. Image fusion algorithm with deformable registration between MR and computed tomography (CT) was developed to solve this issue. However, delivered dose may be different due to volumetric changes during image registration process. In this respect, synthetic CT generated from the MR image would provide more accurate information required for the real time RT. Materials and Methods: We analyzed 1,209 MR images from 16 patients who underwent MR guided RT. Structures were divided into five tissue types, air, lung, fat, soft tissue and bone, according to the Hounsfield unit of deformed CT. Using the deep learning model (U-NET model), synthetic CT images were generated from the MR images acquired during RT. This synthetic CT images were compared to deformed CT generated using the deformable registration. Pixel-to-pixel match was conducted to compare the synthetic and deformed CT images. Results and Discussion: In two test image sets, average pixel match rate per section was more than 70% (67.9 to 80.3% and 60.1 to 79%; synthetic CT pixel/deformed planning CT pixel) and the average pixel match rate in the entire patient image set was 69.8%. Conclusion: The synthetic CT generated from the MR images were comparable to deformed CT, suggesting possible use for real time RT. Deep learning model may further improve match rate of synthetic CT with larger MR imaging data.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2017년도 추계학술대회
• /
• pp.218-220
• /
• 2017

Recently, stitching techniques are used for obtaining wide FOV, e.g., panorama contents, from normal cameras. Despite many proposed algorithms, the no objective quality evaluation method is developed, so the comparison of algorithms are performed only in subjective way. The paper proposes a 'Delaunay-triangulation based objective assessment method' for evaluating the geometric and photometric distortions of stitched or warped images. The reference and target images are segmented by Delaunay-triangulation based on matched points between two images, the average Euclidian distance is used for geometric distortion measure, and the average or histogram of PSNR for photometric measure. We shows preliminary results with several test images and stitching methods for demonstrate the benefits and application.

• 감성과학
• /
• 제14권3호
• /
• pp.385-394
• /
• 2011

본 연구의 목적은 20대 표준체형 여성을 대상으로 하여, 소재에 따른 플레어 스커트의 실제착의와 가상착의에 따른 시각적 이미지를 비교 분석하고, 시각적 이미지와 역학적 특성간의 상관관계를 분석하는 데에 있다. 본 연구는 드레이프 특성이 확연히 다른 5종류의 소재를 사용하였다. 실험에 사용되어진 플레어 스커트의 실제착의와 가상착의의 이미지는 사진으로 제공되었으며, 피설문자는 20대의 의류학 전공의 여성이었다. 자료의 분석은 SPSS Ver.12.0 프로그램을 사용하여 통계 처리하였으며, 연구 문제별로 요인분석, 일원변량분석(One way ANOVA), T 검정(t-test), 던컨테스트(Duncan test)를 실시하였다. 시각적 이미지에 대한 요인분석 결과 '드레이프성', '매력성', '신체 보정성', '부피감', '활동성' 의 총 5 가지 요인이 분석되었다. 시각적 이미지중 '부피감'의 경우 G, 무게, 두께와 같은 역학적 특성들과 밀접한 상관관계를 나타냈으며, 3차원 의복 시뮬레이션과 실제착의간의 시각적 이미지는 소재에 따라 유의한 차이점을 나타냈는데, 실크나 폴리에스터 소재와 면, 린넨, 양모소재간 이미지 차이는 소재의 무게와 두께에 따라 영향을 많이 받는 것으로 나타났다.

• Journal of the Korean Wood Science and Technology
• /
• 제47권4호
• /
• pp.385-392
• /
• 2019

In our previous study, we investigated the use of ensemble models based on LeNet and MiniVGGNet to classify the images of transverse and longitudinal surfaces of five Korean softwoods (cedar, cypress, Korean pine, Korean red pine, and larch). It had accomplished an average F1 score of more than 98%; the classification performance of the longitudinal surface image was still less than that of the transverse surface image. In this study, ensemble methods of two different convolutional neural network models (LeNet3 for smartphone camera images and NIRNet for NIR spectra) were applied to lumber species classification. Experimentally, the best classification performance was obtained by the averaging ensemble method of LeNet3 and NIRNet. The average F1 scores of the individual LeNet3 model and the individual NIRNet model were 91.98% and 85.94%, respectively. By the averaging ensemble method of LeNet3 and NIRNet, an average F1 score was increased to 95.31%.