• The KIPS Transactions:PartB
• /
• v.16B no.5
• /
• pp.395-402
• /
• 2009

This paper proposes an approach of a novel facial expression recognition to deal with different intensities to improve a performance of a facial expression recognition. Various expressions and intensities of each person make an affect to decrease the performance of the facial expression recognition. The effect of different intensities of facial expression has been seldom focused on. In this paper, a face expression template and an expression-intensity distribution model are introduced to recognize different facial expression intensities. These techniques, facial expression template and expression-intensity distribution model contribute to improve the performance of facial expression recognition by describing how the shift between multiple interest points in the vicinity of facial parts and facial parts varies for different facial expressions and its intensities. The proposed method has the distinct advantage that facial expression recognition with different intensities can be very easily performed with a simple calibration on video sequences as well as still images. Experimental results show a robustness that the method can recognize facial expression with weak intensities.

• KIPS Transactions on Computer and Communication Systems
• /
• v.5 no.11
• /
• pp.427-438
• /
• 2016

Car blackboxes record the information and status of driving. Since blackboxes are commonly used in daily life, the usage of video data recorded from blackboxes is increasing for investigating. Investigators use a own analysis tool suitable for their blackbox provided by the manufacturer in order to check the data. But the tools are not enough to use in the digital forensic analysis because they are dependent on a specific model of blackbox and provides ungeneralized functions. Moreover, if the manufacturer is bankrupt, then their own tools can not be obtained also. Therefore, the way data are stored in the blackboxes which are now in the market are investigated and the features and limitations which have blackbox's own analysis tools are checked. And a comprehensive tool for the analysis of blackboxes is designed and implemented as in this paper.

• Geophysics and Geophysical Exploration
• /
• v.23 no.3
• /
• pp.157-167
• /
• 2020

With the recent advancement of computer hardware and the contribution of open source libraries to facilitate access to artificial intelligence technology, the use of machine learning (ML) and deep learning (DL) technologies in various fields of exploration geophysics has increased. In addition, ML researchers have developed complex algorithms to improve the inference accuracy of various tasks such as image, video, voice, and natural language processing, and now they are expanding their interests into the field of automatic machine learning (AutoML). AutoML can be divided into three areas: feature engineering, architecture search, and hyperparameter search. Among them, this paper focuses on hyperparamter search with Bayesian optimization, and applies it to the problem of facies classification using seismic data and well logs. The effectiveness of the Bayesian optimization technique has been demonstrated using Vincent field data by comparing with the results of the random search technique.

• Journal of the Korea Society for Simulation
• /
• v.27 no.2
• /
• pp.1-10
• /
• 2018

This paper presents an autonomous detection and aiming point computation of a battle tank by using RGB images. Maximally stable extremal regions algorithm was implemented to find features of the tank, which are matched with images extracted from streaming video to figure out the region of interest where the tank is present. The median filter was applied to remove noises in the region of interest and decrease camouflage effects of the tank. For the tank segmentation, k-mean clustering was used to autonomously distinguish the tank from its background. Also, both erosion and dilation algorithms of morphology techniques were applied to extract the tank shape without noises and generate the binary image with 1 for the tank and 0 for the background. After that, Sobel's edge detection was used to measure the outline of the tank by which the aiming point at the center of the tank was calculated. For performance measurement, accuracy, precision, recall, and F-measure were analyzed by confusion matrix, resulting in 91.6%, 90.4%, 85.8%, and 88.1%, respectively.

• International Journal of Aeronautical and Space Sciences
• /
• v.10 no.2
• /
• pp.117-124
• /
• 2009

In the environment where various and complicated threat signals exist, RWR (Radar Warning Receiver), which can warn pilot of the existence of threats, has long been a necessary electronic warfare (EW) system to improve survivability of aircraft. The angle of arrival (AOA) information, the most reliable sorting parameter in the RWR, is measured by means of four-quadrant amplitude comparison direction finding (DF) technique. Each of four antennas (usually spiral antenna) of DF unit covers one of four quadrant zones, with 90 degrees apart with nearby antenna. According to the location of antenna installed in helicopter, RWR is subject to signal loss and interference by helicopter body and structures including tail bumper, rotor blade, and so on, causing a difficulty of detecting hostile emitters. In this paper, the performance degradation caused by signal interference by tail rotor blades has been estimated by measuring amplitude video signals into which RWR converts RF signals in case a part of antenna is screened by real tail rotor blade in anechoic chamber. The results show that corruption of pulse amplitude (PA) is main cause of DF error. We have proposed two algorithms for resolving the interference by tail rotor blades as below: First, expand the AOA group range for pulse grouping at the first signal analysis phase. Second, merge each of pulse trains with the other, that signal parameter except PRI and AOA is similar, after the first signal analysis phase. The presented method makes it possible to use RWR by reducing interference caused by blade screening in case antenna is screened by tail rotor blades.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.6
• /
• pp.61-71
• /
• 2004

Recently, various techniques for face detection are studied, but most of them still have problems on processing in real-time. Therefore, in this paper, we propose novel techniques for real-time detection of human faces in sequential images using motion and chroma information. First, background model is used to find a moving area. In this procmoving area. edure, intensity values for reference images are averaged, then skin-color are detected in We use HCr color-space model and adaptive threshold method for detection. Second, binary image labeling is applied to acquire candidate regions for faces. Candidates for mouth and eyes on a face are obtained using differences between green(G) and blue(B), intensity(I) and chroma-red(Cr) value. We also considered distances between eye points and mouth on a face. Experimental results show effectiveness of real-time detection for human faces in sequential images.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.43 no.10 s.352
• /
• pp.45-53
• /
• 2006

In this paper, we proposed a parallel hardware architecture for deblocking filter in K264/AVC. The deblocking filter has high efficiency in H.264/AVC, but it also has high computational complexity. For real time video processing, we chose a two 1-D parallel filter architecture, and tried to reduce memory access using dual-port SRAM. The proposed architecture has been described in Verilog-HDL and synthesized on Hynix 0.25um CMOS Cell Library using Synopsys Design Compiler. The hardware size was about 27.3K logic gates (without On-chip Memory) and the maximum operating frequency was 100Mhz. It consumes 258 clocks to process one macroblock, witch means it can process 47.8 HD1080P(1920pixel* 1080pixel) frames per second. It seems that it can be used for real time H.264/AVC encoding and decoding of various multimedia applications.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.9
• /
• pp.139-147
• /
• 2014

In this paper, we present a novel method for simultaneous detection and tracking of multiple objects using dual-layer particle filtering. The proposed dual-layer particle sampling (DLPS) algorithm consists of parent-particles (PP) in the first layer for detecting multiple objects and child-particles (CP) in the second layer for tracking objects. In the first layer, PPs detect persons using a classifier trained by the intersection kernel support vector machine (IKSVM) at each particle under a randomly selected scale. If a certain PP detects a person, it generates CPs, and makes an object model in the detected object region for tracking the detected object. While PPs that have detected objects generate CPs for tracking, the rest of PPs still move for detecting objects. Experimental results show that the proposed method can automatically detect and track multiple objects, and efficiently reduce the processing time using the sampled particles based on motion distribution in video sequences.

• IEIE Transactions on Smart Processing and Computing
• /
• v.6 no.1
• /
• pp.66-70
• /
• 2017

Radioactive materials are used in medicine, non-destructive testing, and nuclear plants. Source localization is especially important during nuclear decommissioning and decontamination because the actual location of the radioactive source within nuclear waste is often unknown. The coded-aperture imaging technique started with space exploration and moved into X-ray and gamma ray imaging, which have imaging process characteristics similar to each other. In this study, we simulated $21{\times}21$ and $37{\times}37$ coded aperture collimators based on a modified uniformly redundant array (MURA) pattern to make a gamma imaging system that can localize a gamma-ray source. We designed a $21{\times}21$ coded aperture collimator that matches our gamma imaging detector and did feasibility experiments with the coded aperture imaging system. We evaluated the performance of each collimator, from 2 mm to 10 mm thicknesses (at 2 mm intervals) using root mean square error (RMSE) and sensitivity in a simulation. In experimental results, the full width half maximum (FWHM) of the point source was $5.09^{\circ}$ at the center and $4.82^{\circ}$ at the location of the source was $9^{\circ}$. We will continue to improve the decoding algorithm and optimize the collimator for high-energy gamma rays emitted from a nuclear power plant.

• Journal of Digital Contents Society
• /
• v.3 no.1
• /
• pp.1-9
• /
• 2002

This paper is proposed to algorithm for measuring traffic information automatically, for example, volume count, speed and occupancy rate, from CCTV camera images installed on highway, add to function of image detectors which can be collected the traffic information. Recently the method of traffic informations are counted in lane one by one, but this manner is occurred critical errors by occlusion frequently in case of passing larger vehicles(bus, truck etc.) and is impossible to measure in the 8 lanes of highway. In this paper, installed the detection area include with all lanes, traffic informations are collected using tracking algorithm with passing vehicles individually in this detection area, thus possible to detect all of 8 lanes. The experiment have been conducted two different real road scenes for 20 minutes. For the experiments, the images are provided with CCTV camera which was installed at Kiheung Interchange upstream of Kyongbu highway, and video recording images at Chungkye Tunnel. For image processing, images captured by frame-grabber board 30 frames per second, $640{\times}480$ pixels resolution and 256 gray-levels to reduce the total amount of data to be interpreted.