• Journal of the Korea Society of Computer and Information
• /
• v.26 no.1
• /
• pp.45-55
• /
• 2021

This paper has proposed an algorithm that detecting for dense small vehicle in large image efficiently. It is consisted of two Ensemble Deep-Learning Network algorithms based on Coarse to Fine method. The system can detect vehicle exactly on selected sub image. In the Coarse step, it can make Voting Space using the result of various Deep-Learning Network individually. To select sub-region, it makes Voting Map by to combine each Voting Space. In the Fine step, the sub-region selected in the Coarse step is transferred to final Deep-Learning Network. The sub-region can be defined by using dynamic windows. In this paper, pre-defined mapping table has used to define dynamic windows for perspective road image. Identity judgment of vehicle moving on each sub-region is determined by closest center point of bottom of the detected vehicle's box information. And it is tracked by vehicle's box information on the continuous images. The proposed algorithm has evaluated for performance of detection and cost in real time using day and night images captured by CCTV on the road.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.6
• /
• pp.566-572
• /
• 2021

A methodology for the identification and coordinates estimation of air cavities under urban ground or sandy soil using its natural poles and natural resonant frequencies is presented. The potential of this methodology was analyzed. Simulation models of PEC (Perfect Electric Conductor)s with various shapes and dimensions were developed using an EM (Electromagnetic) simulator. The Cauchy method was applied to the obtained EM scattering response of various objects from EM simulation models. The natural poles of objects corresponding to its instinct characterization were then extracted. Thus, a library of poles can be generated using their natural poles. The generated library of poles provided the possibility of identifying a target by comparing them with the computed natural poles from a target. The simulation models were made assuming that there is an air cavity under urban ground or sandy soil. The response of the desired target was extracted from the electromagnetic wave scattering data from its simulation model. The coordinates of the target were estimated using the time delay of the impulse response (peak of the impulse response) in the time domain. The MP (Matrix Pencil) method was applied to extract the natural poles of a target. Finally, a 0.2-m-diameter spherical air cavity underground could be estimated by comparing both the pole library of the objects and the calculated natural poles and the natural resonant frequency of the target. The computed location (depth) of a target showed an accuracy of approximately 84 to 93%.

• Pediatric Gastroenterology, Hepatology & Nutrition
• /
• v.10 no.2
• /
• pp.157-165
• /
• 2007

Purpose: Ingested foreign bodies are removed by endoscopy, surgery or spontaneous passage, however, the decision of therapeutic modality chosen depends on the type, size, shape and location of the ingested foreign bodies. Therefore, this study was conducted to investigate the rate and characteristics of foreign bodies that were passed spontaneously out of the intestine. Methods: One hundred and sixty patients who visited Samsung Medical Center for treatment of gastrointestinal foreign bodies between January 2001 to July 2007 were enrolled in this study. Related data was colleted by reviewing the medical records of patients with proven foreign bodies retrospectively, as well as by conducting phone interviews with the parents of the patients. All cases were classified based on the nature and location of the ingested foreign body, as well as whether it was treated by spontaneous passage. Results: Of the 160 cases involving the passage of foreign bodies in children were included in this study (95 boys and 65 girls), endoscopic removals, operative removals or spontaneous passages were conducted in 80, 3 and 77 patients, respectively. The spontaneous passage rates for each type of object were as follows; coins (36.5%), bead and baduk stones (83.3%), long and sharp materials (52.6%), magnets (69.2%) and disc batteries (50.0%). In cases involving round-shaped foreign bodies, such as coins, the diverse spontaneous passage rates were more diverse. When foreign bodies were stuck below esophagus, similar sized coins and baduk stones had spontaneous passage rates greater than 80% regardless of age. Conclusion: It is better for clinicians to wait for spontaneous passage to occur in cases involving coins or round-shaped foreign bodies that are located at or below the stomach.

• The Korean Journal of Vision Science
• /
• v.20 no.4
• /
• pp.469-482
• /
• 2018

Purpose : We investigated the change of spherical and comma aberrations after wearing aspheric soft contact lens (ASCL) in young myopes. Methods : Fifty young myopes ($23.15{\pm}1.70years$, spherical equivalent: $-2.90{\pm}1.75D$) were recruited and refractive errors were corrected using ASCL (Biotrue, Bausch+Lomb, USA). High order aberrations were measured in the 4 mm pupil size using the wavefront analyze and pupil sizes were measured with a pupillometer at the modes of scotopic condition (light off) at 3.5 m in the 100 lx illuminance condition. Results : Spherical aberrations and coma aberration of the 20s myopes were $0.026{\pm}0.031{\mu}m$ and $0.078{\pm}0.039{\mu}m$ respectively, and $0.019{\pm}0.026{\mu}m$ and $0.082{\pm}0.038{\mu}m$ after ASCL wear that spherical aberration was decreased and coma aberration was increased. However, spherical aberration was decreased in the 68% of the subject have positive spherical aberration, and increased in the 11% of the subject have negative spherical aberration. Coma aberration was increased in the 53% of the subject, did not change in the 19% of the subjects, and decreased in the 28% of the subject. Spherical aberration was not different with the refractive errors in low and moderate myopies, however, coma aberrations was higher in the higher myopes. Conclusion : In a scotopic condition without accommodation stimuli, spherical aberration is decreased after wearing ASCL, however in the subject have negative spherical aberration spherical aberration could be increased, and which is thought to be the influence of contact lens design and pupil size.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.23 no.2
• /
• pp.270-273
• /
• 2012

A plastic bolt shape UHF RFID tag has been developed for a live tree. The UHF tag is designed and installed into the head part of the bolt, inserted into a tree for management of tree. If the tag antenna is installed near the high dielectric constant material, the impedance of the tag antenna will be changed, and the tag does not work. Therefore, the dielectric constants of wood and plastic bolt are considered for tag antenna design. The input reflection coefficient characteristics and the reading range patterns are measured and compared. This UHF RFID tag can be applied into a live tree, and the status and location of tree can be controlled with the RFID tag. This developed UHF tag can be applied to any applications and objects using a bolt.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.1
• /
• pp.418-424
• /
• 2014

In this paper, we propose a boundary input method for video surveillance systems. Since intrusion of a moving object is decided by comparition of its position and the surveillance boundary, the boundary input method is a basic function in video surveillance. Previous methods are difficult to adapt to the change of surveillance environments such as the size of surveillance area, the number of cameras, and the position of cameras because those build up the surveillance boundary using the captured image in the center of each camera. In our approach, the whole surveillance boundary is once defined in the form of polygon based on the satellite map and transformed into each camera environment. Its characteristics is that the boundary input is independent from the surveillance environment. Given the position of a moving object, the time complexity of its intrusion detection shows O(n), where n is the number of polygon vertices. To verify our method, we implemented a 3D simulation and assured that the input boundary can be reused in each camera without any redefinition.

• Journal of Digital Convergence
• /
• v.12 no.1
• /
• pp.519-524
• /
• 2014

This research is the study of raindrop rendering. In case of rendering for raindrop in existing games, it is used on sprites images or roughly raindrops images using texture rendering. These methods are similar to the shape and size of all rendered raindrop. That's why players are limitations to provide a sense of reality. To overcome this limitation, this paper proposes a method for generating raindrop considering surface tension and contac angle, the amount of water, implements the raindrop using Unity3D engine. To demonstrate the usefulness of this paper, this paper shows the generated raindrop in accordance with the change in the area and pulling force in surface tension formula. This paper can help to provide the actuality in game in case of rendering the raindrop.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.41 no.9
• /
• pp.49-56
• /
• 2004

A new readout circuit(ROIC) for linear HgCdTe 64${\times}$2 two-color Infrared (IR) detector is described. This circuit is based on the buffered direct injection(BDI) technology with high injection efficiency. By using saturation current isolation circuit, the proposed ROIC removed the problems that LWIR(Long Wavelength InfraRed) signal distort when MWIR(Middle Wavelength InfraRed) signal saturates so that new ROIC has larger measurable temperature range about 120k than that of previous circuit and it is also tolerant for dead pixel in MWIR detector. The designed circuit was fabricated using 0.6um 2-poly 3-metal CMOS process. We measured that the designed circuit outputs MWIR signal and LWIR signal simultaneously and saturation current isolationcircuit also operates well. Next, measured noise was about 53uV at room temperature and it can be assumed that designed circuit can satisfy nearly 95% BLIP condition at 77K.

• Journal of the Korea Society of Computer and Information
• /
• v.6 no.2
• /
• pp.65-69
• /
• 2001

Edge detection is the most important process that belongs to the first step in image recognition or vision system and can determine the efficiency valuation. The edge detection with color images is very difficult. because color images have lots of information that contain not only general information representing shape, brightness and so on but also that representing colors. In this paper, we propose architecture of universalized Color Morphological Pyramids(CMP) which is able to give effective edge detection. Image pyramid architecture is a successive image sequence whose area ratio 2$\^$-1/(ι= 1, 2, . . . ,N) after filtering and subsampling of input image. In this technique, noise removed by sequential filtering and resolution is degraded by downsampling using CMP in various color spaces. After that, new level images are constructed that apply formula using distance of neighbor vectors in close level images and detection its image.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.42 no.5 s.305
• /
• pp.19-26
• /
• 2005

In this paper, a new method for obtaining three-dimensional shape of an object by measuring relative blur between images using wavelet analysis has been described. Most of the previous methods use inverse filtering to determine the measure of defocus. These methods suffer from some fundamental problems like inaccuracies in finding the frequency domain representation, windowing effects, and border effects. Besides these deficiencies, a filter, such as Laplacian of Gaussian, that produces an aggregate estimate of defocus for an unknown texture, can not lead to accurate depth estimates because of the non-stationary nature of images. We propose a new depth from defocus (DFD) method using wavelet analysis that is capable of performing both the local analysis and the windowing technique with variable-sized regions for non-stationary images with complex textural properties. We show that normalized image ratio of wavelet power by Parseval's theorem is closely related to blur parameter and depth. Experimental results have been presented demonstrating that our DFD method is faster in speed and gives more precise shape estimates than previous DFD techniques for both synthetic and real scenes.