• Korean Journal of Remote Sensing
• /
• v.37 no.3
• /
• pp.475-491
• /
• 2021

With the recent increase in available high-resolution (< ~1 m) satellite SAR images, the demand for precise registration of SAR images is increasing in various fields including change detection. The registration between high-resolution SAR images acquired in different look angle is difficult due to speckle noise and geometric distortion caused by the characteristics of SAR images. In this study, registration is performed in two stages, coarse and fine, using the x-band SAR data imaged at staring spotlight mode of TerraSAR-X. For the coarse registration, a method combining the adaptive sampling method and SAR-SIFT (Scale Invariant Feature Transform) is applied, and three rigid methods (NCC: Normalized Cross Correlation, Phase Congruency-NCC, MI: Mutual Information) and one non-rigid (Gefolki: Geoscience extended Flow Optical Flow Lucas-Kanade Iterative), for the fine registration stage, was performed for performance comparison. The results were compared by using RMSE (Root Mean Square Error) and FSIM (Feature Similarity) index, and all rigid models showed poor results in all image combinations. It is confirmed that the rigid models have a large registration error in the rugged terrain area. As a result of applying the Gefolki algorithm, it was confirmed that the RMSE of Gefolki showed the best result as a 1~3 pixels, and the FSIM index also obtained a higher value than 0.02~0.03 compared to other rigid methods. It was confirmed that the mis-registration due to terrain effect could be sufficiently reduced by the Gefolki algorithm.

• Journal of Astronomy and Space Sciences
• /
• v.22 no.4
• /
• pp.393-408
• /
• 2005

We developed and compared two automatic algorithms for moving object detections in the YSTAR-NEOPAT sky survey program. One method, called starlist comparison method, is to identify moving object candidates by comparing the photometry data tables from successive images. Another method, called image subtraction method, is to identify the candidates by subtracting one image from another which isolates sources moving against background stars. The efficiency and accuracy of these algorithms have been tested using actual survey data from the YSTAR-NEOPAT telescope system. For the detected candidates, we performed eyeball inspection of animated images to confirm validity of asteroid detections. Main conclusions include followings. First, the optical distortion in the YSTAR-NEOPAT wide-field images can be properly corrected by comparison with USNO-B1.0 catalog and the astrometric accuracy can be preserved at around 1.5 arcsec. Secondly, image subtraction provides more robust and accurate detection of moving objects. For two different thresholds of 2.0 and $4.0\sigma$, image subtraction method uncovered 34 and 12 candidates and most of them are confirmed to be real. Starlist comparison method detected many more candidates, 60 and 6 for each threshold level, but nearly half of them turned out to be false detections.

• Journal of the Korean Society of Radiology
• /
• v.83 no.2
• /
• pp.344-359
• /
• 2022

Purpose To develop a denoising convolutional neural network-based image processing technique and investigate its efficacy in diagnosing breast cancer using low-dose mammography imaging. Materials and Methods A total of 6 breast radiologists were included in this prospective study. All radiologists independently evaluated low-dose images for lesion detection and rated them for diagnostic quality using a qualitative scale. After application of the denoising network, the same radiologists evaluated lesion detectability and image quality. For clinical application, a consensus on lesion type and localization on preoperative mammographic examinations of breast cancer patients was reached after discussion. Thereafter, coded low-dose, reconstructed full-dose, and full-dose images were presented and assessed in a random order. Results Lesions on 40% reconstructed full-dose images were better perceived when compared with low-dose images of mastectomy specimens as a reference. In clinical application, as compared to 40% reconstructed images, higher values were given on full-dose images for resolution (p < 0.001); diagnostic quality for calcifications (p < 0.001); and for masses, asymmetry, or architectural distortion (p = 0.037). The 40% reconstructed images showed comparable values to 100% full-dose images for overall quality (p = 0.547), lesion visibility (p = 0.120), and contrast (p = 0.083), without significant differences. Conclusion Effective denoising and image reconstruction processing techniques can enable breast cancer diagnosis with substantial radiation dose reduction.

• Radiation Oncology Journal
• /
• v.25 no.4
• /
• pp.268-277
• /
• 2007

Purpose: Respiratory motion is a considerable inhibiting factor for precise treatment with stereotactic radiosurgery using the CyberKnife (CK). In this study, we developed a moving phantom to simulate three-dimensional breathing movement and investigated the distortion of dose profiles between the use of a moving phantom and a static phantom. Materials and Methods: The phantom consisted of four pieces of polyethylene; two sheets of Gafchromic film were inserted for dosimetry. Treatment was planned to deliver 30 Gy to virtual tumors of 20, 30, 40, and 50 mm diameters using 104 beams and a single center mode. A specially designed robot produced three-dimensional motion in the right-left, anterior-posterior, and craniocaudal directions of 5, 10 and 20 mm, respectively. Using the optical density of the films as a function of dose, the dose profiles of both static and moving phantoms were measured. Results: The prescribed isodose to cover the virtual tumors on the static phantom were 80% for 20 mm, 84% for 30 mm, 83% for 40 mm and 80% for 50 mm tumors. However, to compensate for the respiratory motion, the minimum isodose levels to cover the moving target were 70% for the $30{\sim}50$ mm diameter tumors and 60% for a 20 mm tumor. For the 20 mm tumor, the gaps between the isodose curves for the static and moving phantoms were 3.2, 3.3, 3.5 and 1.1 mm for the cranial, caudal, right, and left direction, respectively. In the case of the 30 mm tumor, the gaps were 3.9, 4.2, 2.8, 0 mm, respectively. In the case of the 40 mm tumor, the gaps were 4.0, 4.8, 1.1, and 0 mm, respectively. In the case of the 50 mm diameter tumor, the gaps were 3.9, 3.9, 0 and 0 mm, respectively. Conclusion: For a tumor of a 20 mm diameter, the 80% isodose curve can be planned to cover the tumor; a 60% isodose curve will have to be chosen due to the tumor motion. The gap between these 80% and 60% curves is 5 mm. In tumors with diameters of 30, 40 and 50 mm, the whole tumor will be covered if an isodose curve of about 70% is selected, equivalent of placing a respiratory margin of below 5 mm. It was confirmed that during CK treatment for a moving tumor, the range of distortion produced by motion was less than the range of motion itself.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.13 no.3
• /
• pp.10-16
• /
• 2009

Purpose: As the number of patients has increased since the installation of a PET/CT, we are now examining about 2500-3000 annually. We have realized that if we properly adjust a pitch under the same condition of a CT during a PET/CT exam, radiation quantity that reaches the patient can change. In order to reduce the exposure dose of a patient, the research examines a method of reducing the exposure dose of a patient by controlling the pitch during a PET/CT exam, viewing whether the adjustment of the pitch influences CT image and PET SUV. Methods: The equipment used is a Biograph Positron Emission Tomography (PET) Scanner (CT type: TRCT-240-130 (WCT-240-130)) of Siemens company. For the evaluation of exposure dose of a patient, we measured radiation quantities using a PTW-DIADOS 11003/1383, which is a CT radiation measurement instrument used by Siemens. We measured and analyzed the space resolutions of CT images caused by the change of pitches using an AAPM Standard Phantom in order to see how the adjustment of pitches influenced the CT images. In addition, in order to obtain SUVs caused by each change of pitches using a PET source made with a solid radioactive cylinder phantom, we confirmed whether the SUVs changed in the PET/CT images by calculating the SUVs of the fusion images caused by the change of pitches after obtaining CT and PET images and finishing the test. Results: 2slice CT scanner showed that radiation quantities largely dropped when pitches ranged from 0.7 to 1.3 and that the reduction of radiation quantities were smaller when pitches ranged from 1.5 to 1.9. That is, we found that the bigger pitch values are the smaller the radiation quantities of a patient are. Moreover, we realized that there is no change of SUVs caused by the increase of pitches and that pitch values do not influence PET SUVs and the quality of CT images. It is judged that using 1.5 as a pitch value contributes to the reduction of exposure dose of a patient as long as there is no problem in the quality of an image. Conclusions: When seeing the result of the research, hospital using a PET/CT should make an effort to reduce the exposure dose of a patient seeking pitch values appropriate for their hospital within the range in which there is no image distortion and PET SUVs are not influenced from pitches. We think that the research can apply to all multi-detectors having a CT scanner and that such a research will be needed for other equipments in the future.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.14 no.1
• /
• pp.3-7
• /
• 2010

Purpose: The usefulness of Positron Emission Tomography (PET) images in diagnosis, staging, recurrent and treatment response evaluation has already been known. However, tumors which are small size, located in lower lobe of lung or upper lobe of liver are shown misalignment, distortion and different Standard Uptake Value (SUV) by respiration in PET images. Therefore, if radiotherapy based on normal respiration, it may cause low treatment response or more side effects because targets which had to treat, out of treat range or over dose to normal tissue. The purpose of this study is to evaluate attenuation-correction with Average CT (ACT) for more accuracy SUV measurement and minimize artifact by respiration. Materials and Methods: 13 patients, who had tumors which are around the diaphragm, underwent ACT scan after Helical CT (HCT) scan with PET/CT (Discovery DSTE 8; GE Healthcare). We quantified the differences between attenuation corrected image with HCT and attenuation corrected image with ACT in artifact size and maximum SUV ($SUV_{max}$). Artifacts were evaluated by measurement of the curved photogenic area in the lower thorax of the PET images for all patients. $SUV_{max}$ was measured separately at the primary tumors. Analysis program was Advantage Workstation v4.3 (GE Healthcare). Patients were injected with 7.4 MBq (0.2 $mC_i$) per kg of $^{18}F$-FDG and scanned 1 hour after injection. The PET acquisition was 3 minute per bed. Results: Significantly lower artifact were observed in PET/ACT images than in PET/HCT images (below-thoracic artifacts caused by under corrected $1.5{\pm}3.5$ cm vs. $13.4{\pm}4.2$ cm). Significantly higher $SUV_{max}$ were noted in PET/ACT images than in PET/HCT images in the primary tumor. Compared with PET/HCT images, $SUV_{max}$ in PET/ACT images were higher by $5.3{\pm}3.9%$ (mean value) tumor. The highest difference was observed in Lower lobe of lung (7.7 to 8.7; 13%). Conclusion: Due to its significantly reduced artifacts in lower thoracic, attenuation corrected image with ACT images provided more reliable $SUV_{max}$ and may be helpful in monitoring treatment response. Moreover, ACT can separate upper lobe of liver and lower lobe of lung, it may be helpful in interpretation. ACT will be clinically useful, considering increased dose caused by ACT scan and adapt.

• Journal of the Korean Society of Radiology
• /
• v.10 no.4
• /
• pp.255-261
• /
• 2016

In this study, among various factors having influence on SUV, we intended to compare and analyze the change of SUV using CT(4 type) and MRI(3 type) contrast agents which are commercialized now. We used Discovery 690 PET/CT(GE) and NEMA NU2 - 1994 PET phantom as experimental equipment. We have conducted a study as follows; first, we filled distilled water to phantom about two-thirds and injected radioisotope(18F-FDG 37 MBq), contrast agent. Second, we mixed CT contrast agent with distilled water and MRI contrast agent with that water separately. And then, we stirred the fluid and filled distilled water fully not to make air bubble. In emission scan, we had 15minutes scanning time after 40 minutes mixing contrast agent with distilled water. In transmission scan, we used CT scanning and its measurement conditions were tube voltage 120 kVp, tube current 40 mA, rotation time 0.5 sec, slice thickness 3.27 mm, DFOV 30 cm. Analyzing results, we set up some ROIs in 10th, 15th, 20th, 25th, 30th slice and measured SUVmean, SUVmax. Consequently, all images mixed 3 types of MRI contrast agent with distilled water have high SUVmean as compared with pure FDG image but there was no statistical significance. In SUVmax, they have high score and there was statistical significance. And other 4 images mixed 4 types of CT contrast agent with distilled water have significance in both SUVmean and SUVmax. Attenuation correction in PET/CT has been executed through various methods to make high quality image. But we figured out that using CT and MRI contrast agents before PET/CT scanning could make distortion of image and decrease diagnostic value. In that reason, we have to sort out the priority of examination in hospital not to disturb other examination's results. Through this process, we will be able to give superior medical service to our customers.

• Journal of the Korean Society of Costume
• /
• v.32
• /
• pp.293-312
• /
• 1997

Deconstructionism is a philosophical stream in the latter half of 20'th century which goes against western metaphysics and tries to deconstruct the dualism e.g. eastern/west-ern man/woman. Its main concepts are 'indi-vidual' 'other' 'difference' 'restoration of the repressed' 'decentralization' 'Today it shows strong influences in literature art, and other social fields. This study investigated inner meanings and exterior forms represented in clothing on the base of Jacques Derrida's theory in order to re-search modes of deconstruction in is. And it compared clothing with architecture among other genres of art in order to discover similarities between through and modes of art in a certain period. For illuminating concept of deconstruction I referred to the literatures of preceding studies and for deconstruction's characteristics in both clothing and architec-ture I referred work's collection book $\ulcorner$collec-tions$\lrcorner$ and other related books dealing from 1960's to this point. As a rsult there are four modes of decons-truction in both clothing and architecture as follows; 1) Differenance which is a concept of the dif-ference in time and space as being trace. 2) In termeaning of meanings which is not present in reality but re-interpreted a new in the future. 3) Interxtuality in which a texte is not alone but with others 4) Dis.De phenomenon in which distortion and fragmentation forms occur as the result of the denial of perfection and reson. There are characteristics of modes of de-construction in clothing as follows; 1) Differance; R.Gigli, P. Rabanne. G.Bersace, V.Westwood are representatives of the cloth-ing which is combined with the past the pres-ent and the future or is re-viewed as the clothing of the past in the present point of view. J. Watanabe R. Dawakubo I miyake are representatives of the clothing which is out-date but re-viewed in the sense of recollec-tion and re-usage. 2) Intermeaning of meanings: R. kawakubo I, Miyake Y,Yamamoto are representatives of the clothing which is incomplete but re-interpreted in the future and appears to be new-fashioned according to ways of bnding wearing throwing on and tying J.C. Castel-bajac K.Hamnett S.Sprouse are representa-tives of the clothing which is made up of ab -stract forms scribbling or symbolic letters which appears to be different according to view-point of observers. 3) Intertextuality ;J.P.Gaultier K.Hamnett, Comme des Gar ons are representatives of intertextuality of gender which avoids the 여-ality of man/woman J.P Gaultier G,Versace J. galliano are representatives of the intertex-tuality of time place and occasion which mixes temporality extensity and purposiveness. J.P Gaultier G,Versace are representatives of the intertextuality of coordination which combines items different in image purpose and use. P.Rabanne A,Courr ges R,Gernreic I,Miyake are representatives of the intertextuality of mat-ter which uses heterogeneous matter different from cloth. 4) Dis.De-phenomenon: R, Kawakubo I.Mi-yake J.P.Gaultier are representatives of the de-composition which discloses distortion and exag-geration of form through destructing the estab-lished way of construction J.Galliano R.Kawa-kubo Devota Y Lomba arte representatives of the decentring which restores the repressed and the alienated on the one hane and shows front-centrality on the other. Comme des Gar ones is representative of the discontinuity because of which right and left up and down are not in har-mony with one another. J,Galliano J.P,Gaultier T,Mugler are representatives of the disruption by way of which one makes one's body exposed through intentional slashes or holes. As a consequence deconstructionism enabled us to investigate similarities between through of deconstruction and modes of art interms of diffrance Intermeaning of meanings intertextuality and Dis De-phenomenon. And we found that deconstruction was a phase of development in that it as a all-comprising and multiple concept tries to pursue the new through deconstruction.

• Journal of Broadcast Engineering
• /
• v.18 no.1
• /
• pp.66-76
• /
• 2013

Although there has been much advance in the development of 3D displays of various purpose, 3D contents are not yet so used as expected in those displays. One well-known obstacle in the enjoyment of 3D contents is visual fatigue, but another major issue is image distortion of 3D contents. In the previous research, Shin, Li, & Kim (2012) reported systematic linear relationship between screen disparity and size perception of a simple object whose retinal size was constant across different disparities. In this research, we intended to generalize the previous finding by using various 3D stimuli in the test of the relationship between screen disparity and size perception of those stimuli. Consistent with previous findings, our data indicated that size perception linearly changes as a function of screen disparity and the linearity was observed in all stimuli types we used in this research. We described the empirical relationship between screen disparity and size perception in the form of prediction function for size perception in which visual angle is the predictor. This function will be very useful in the creation of 3D contents as one can make reasonable predictions on the to-be-perceived size of an object being filmed using screen disparity of their camera setting.

• Korean Journal of Optics and Photonics
• /
• v.21 no.4
• /
• pp.151-160
• /
• 2010

When the optical image stabilization is implemented by moving one of the lens groups in a zoom system, decentration should be considered in the optical design process. Although it is partially possible to calibrate optical performances in an optical system with non-symmetrical elements by using a lot of commercial software, the results of calibrating longitudinal aberrations have some calibration errors because of the lack of precise consideration of decentered optical systems. In particular, the amount of distortion in paraxial ray tracing is different from the experimental value because paraxial ray tracing in the optical system is not useful. In this paper, in order to solve this problem being from various commercial lens design software, the set of equations of paraxial ray tracing in a zoom lens system with the non-symmetrical elements like decentration or tilt are theoretically induced. Then, the methods to calibrate the equations of longitudinal aberrations by using these equations in a non-symmetrical optical system are presented. The method of calibrating longitudinal aberrations can in practice be used to correct hand shaking effects in a zoom lens system.