• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.5
• /
• pp.217-228
• /
• 2023

Video instance segmentation is an intelligent visual task with high complexity because it not only requires object instance segmentation for each image frame constituting a video, but also requires accurate tracking of instances throughout the frame sequence of the video. In special, human instance segmentation in drama videos has an unique characteristic that requires accurate tracking of several main characters interacting in various places and times. Also, it is also characterized by a kind of the class imbalance problem because there is a significant difference between the frequency of main characters and that of supporting or auxiliary characters in drama videos. In this paper, we introduce a new human instance datatset called MHIS, which is built upon drama videos, Miseang, and then propose a novel video data augmentation method, CDVA, in order to overcome the data imbalance problem between character classes. Different from the previous video data augmentation methods, the proposed CDVA generates more realistic augmented videos by deciding the optimal location within the background clip for a target human instance to be inserted with taking rich spatio-temporal context embedded in videos into account. Therefore, the proposed augmentation method, CDVA, can improve the performance of a deep neural network model for video instance segmentation. Conducting both quantitative and qualitative experiments using the MHIS dataset, we prove the usefulness and effectiveness of the proposed video data augmentation method.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.23 no.4
• /
• pp.415-420
• /
• 2017

In this paper, an algorithm for object counting via a conveyor system using machine vision is suggested. Object counting systems using image processing have been applied in a variety of industries for such purposes as measuring floating populations and traffic volume, etc. The methods of object counting mainly used involve template matching and machine learning for detecting and tracking. However, operational time for these methods should be short for detecting objects on quickly moving conveyor belts. To provide this characteristic, this algorithm for image processing is a region-based method. In this experiment, we counted young abalone shells that are similar in shape, size and color. We applied a characteristic conveyor system that operated in one direction. It obtained information on objects in the region of interest by comparing a second frame that continuously changed according to the information obtained with reference to objects in the first region. Objects were counted if the information between the first and second images matched. This count was exact when young shells were evenly spaced without overlap and missed objects were calculated using size information when objects moved without extra space. The proposed algorithm can be applied for various object counting controls on conveyor systems.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.47 no.11
• /
• pp.13-22
• /
• 2010

This paper proposed the dual-loops multiphase DLL based mixed VCO/VCDL for a high frequency phase noise suppression of the input clock and the multiple frequencies generation with a precise duty cycle. In the proposed architecture, the dual-loops DLL uses the dual input differential buffer based nMOS source-coupled pairs at the input stage of the mixed VCO/VCDL. This can easily convert the input and output phase transfer of the conventional DLL with bypass pass filter characteristic to the input and output phase transfer of PLL with low pass filter characteristic for the high frequency input phase noise suppression. Also, the proposed DLL can correct the duty-cycle error of multiple frequencies by using only the duty-cycle correction circuits and the phase tracking loop without additional correction controlled loop. At the simulation result with $0.18{\mu}m$ CMOS technology, the output phase noise of the proposed DLL is improved under -13dB for 1GHz input clock with 800MHz input phase noise. Also, at 1GHz operating frequency with 40%~60% duty-cycle error, the duty-cycle error of the multiple frequencies is corrected under $50{\pm}1%$ at 2GHz the input clock.

• Proceedings of the Korean Society of Marine Engineers Conference
• /
• 2011.06a
• /
• pp.235-235
• /
• 2011

Lau basin of south Pacific, as an active back arc basin, is promising area bearing seafloor massive hydrothermal deposit that is located in a subduction zone between the Pacific ocean plate and Indo-Australian continental plate. Korea Ocean Research and Development Institute tracked from 2004 to 2006 the hydrothermal activity to the extension of the northeast Lau Basin, targeting seamount. hydrothermal activity by tracking was found hydrothermal evidences. In this study, Marine seismic survey was carried out in the Lau basin seamount of the possibility of hydrothermal deposit. In particular, Marine magnetic survey and seismic survey was carried out at the same time in TA-12 seamount and noise characteristics were found in the seamount. the main process of data processing is Bandpass filter, FK filter, Deconvolution for noise attenuation such backscatter and multiple reflections. the migration is performed to compensate for reflection points followed by seamount of a slope. In this study, bedrock and upper strata could be identified and in the Future, the comparative method with Multi Beam Echo Sounder(MBES) are likely to derive the correct velocity model, the marine magnetic survey results should be considered.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.3
• /
• pp.569-575
• /
• 2020

Recently, research on renewable energy technologies has come into the spotlight due to rising concerns over the depletion of fossil fuels and greenhouse gas emissions. Demand for portable electronic and wearable devices is increasing, and electronic devices are becoming smaller. Energy harvesting is a technology for overcoming limitations such as battery size and usage time. In this paper, the V-I characteristic curve and internal resistance of thermal electric devices were analyzed, and MPPT control methods were compared. The Perturbation and Observation (P&O) control method is economically inefficient because two sensors are required to measure the voltage and current of a Thermoelectric Generator(TEG). Therefore, this paper proposes a new MPPT control method that tracks MPP using only one sensor for the regulation of the output voltage. The proposed MPPT control method uses the relationship between the output voltage of the load and the duty ratio. Control is done by periodically sampling the output voltage of the DC-DC converter to increase or decrease the duty ratio to find the optimal duty ratio and maintain the MPP. A DC-DC converter was designed using a cascaded boost-buck converter, which has a two-switch topology. The proposed MPPT control method was verified by simulations using PSIM, and the results show that a voltage, current, and power of V=4.2 V, I=2.5 A, and P=10.5 W were obtained at the MPP from the V-I characteristic curve of the TEG.

• Korean Journal of Optics and Photonics
• /
• v.31 no.2
• /
• pp.59-70
• /
• 2020

We discuss a modified numerical model based on the Monte Carlo radiative transfer (MCRT) method, i.e., the MCRT matrix method, for the analysis of atmospheric scattering effects in three-dimensional flash LIDAR systems. Based on the MCRT method, the radiative transfer function for a LIDAR signal is constructed in a form of a matrix, which corresponds to the characteristic response. Exploiting the superposition and convolution of the characteristic response matrices under the paraxial approximation, an extended computer simulation model of an overall flash LIDAR system is developed. The MCRT matrix method substantially reduces the number of tracking signals, which may grow excessively in the case of conventional Monte Carlo methods. Consequently, it can readily yield fast acquisition of the signal response under various scattering conditions and LIDAR-system configurations. Using the computational model based on the MCRT matrix method, we carry out numerical simulations of a three-dimensional flash LIDAR system operating under different atmospheric conditions, varying the scattering coefficient in terms of visible distance. We numerically analyze various phenomena caused by scattering effects in this system, such as degradation of the signal-to-noise ratio, glitches, and spatiotemporal spread and time delay of the LIDAR signals. The MCRT matrix method is expected to be very effective in analyzing a variety of LIDAR systems, including flash LIDAR systems for autonomous driving.

• Proceedings of KOSOMES biannual meeting
• /
• 2006.11a
• /
• pp.277-282
• /
• 2006

The purpose in having a control surface on a ship is to control the motion of the ship. The control surface may be composed entirely of a single movable surface or of a combination of fixed and movable portion A control surface has one sole function to perform in meeting its purpose, and that is to develop a control force in consequence of its orientation and movement relative to the water. The forces and moments generated as a result of this rotation and angle of attack then determine the maneuvering characteristics of the ship. In this paper the study of flapped rudder's 2-dimensional section was accomplished. Model tests had been carried out with different angles of attack of a main foil and flap's deflection angles to predict the performance of the flapped rudder and the 2 frame particle tracking method had been used to obtain the velocity distribution in the flow field $Re=2.8\times10^4$ had been used during the whole experiments and measured results had been compared with each other.

• Journal of Satellite, Information and Communications
• /
• v.7 no.3
• /
• pp.135-139
• /
• 2012

A digital controller can simply realize a complex operation algorithm and power control process which can not be applied by an analog circuit for a solar array regulator(SAR). The digital resistive control(DRC) makes an equivalent input impedance of the SAR be resistive characteristic. The resistance of the solar array varies largely in a voltage source region and slightly in a current source region. Therefore when the solar array regulator is controlled by the DRC, the Advanced Incremental Conductance MPPT Algorithm with a Variable Step Size(AIC-MPPT-VSS) is suitable. The AIC-MPPT-VSS, however, using small signal resistance and large signal resistance of the solar array can not limit the absolute value of the solar array power. In this paper, the solar array power limiter is suggested and the BUCK-BOOST type SAR which is fully controlled by the digital controller is verified by simulation.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.06a
• /
• pp.678-681
• /
• 2007

In this study, an implementation of a system for measuring more accurate blood pressure by non-invasive methods of oscillometric was performed. The system were composed of pressure control, signal measurement, blood pressure signal processing units and wireless sensor network parts. For verify the validity of the system, tests of characteristics evaluations for pressure measurement unit, extraction of characteristic ratios for blood pressure estimation, blood pressure tracking by oscillometric method were performed. A group of five adult male was selected for the clinical test of the implemented system. The results of the oscillometric method in comparison with auscultatory method are that the maximum ratios of PAD of average, systolic and diastolic arterial pressure are 1.38%, 1.63% and 2.97% with SEP of 5.00, 3.72 and 4.34.

• Journal of Korea Multimedia Society
• /
• v.3 no.5
• /
• pp.433-444
• /
• 2000

The recognition of car plate was investigated by means of the enhanced fuzzy ART algorithm. The morphological information of horizontal and vertical edges was used to extract a plate area from a car image. In addition, the contour tracking algorithm by utilizing the SOFM was applied to extract the specific area which includes characters from an extracted plate area. The extracted characteristic area was recognized by using the enhanced fuzzy ART algorithm. In this study we propose the novel fuzzy ART algorithm different from the conventional fuzzy ART algorithm by the dynamical establishment of the vigilance threshold which shows a tolerance limit of unbalance between voluntary and saved patterns for clustering. The extraction rate obtained by using the morphological information of horizontal and vertical edges showed better results than that from the color information of RGB and HSI. Furthermore, the recognition rate of the enhanced fuzzy ART algorithm was improved much more than that of the conventional fuzzy ART and SOFM algorithms.