• Transactions of the Korean Society of Automotive Engineers
• /
• v.16 no.6
• /
• pp.134-140
• /
• 2008

Preceding vehicle detection is a crucial issue for driver assistance system as well as for autonomous vehicle guidance function and it has to be performed with high reliability to avoid any potential collision. The vision-based preceded vehicle detection systems are regarded promising for this purpose because they require little infrastructure on a highway. However, the feasibility of these systems in passenger car requires accurate and robust sensing performance. In this paper, an preceded vehicle detection system is developed using stereo vision sensors. This system utilizes feature matching, epipoplar constraint and feature aggregation in order to robustly detect the initial corresponding pairs. After the initial detection, the system executes the tracking algorithm for the preceded vehicles including a leading vehicle. Then, the position parameters of the preceded vehicles or leading vehicles can be obtained. The proposed preceded vehicle detection system is implemented on a passenger car and its performances is verified experimentally.

• Journal of Sensor Science and Technology
• /
• v.22 no.2
• /
• pp.156-161
• /
• 2013

This study was aimed to implement the electroglottography (EGG) system for analyzing fundamental frequency of the phonation. EGG was recorded from the conductance between ring electrodes attached to the neck skin area near thyroid cartilage with high frequency carrier electric signals during vocalization, and voice signal was recorded with microphone simultaneously. EGG and voice signals were transmitted to the audio port in PC and recorded with stereo sound recording program. From the digitized data, several parameters such as pitch, jitter, shimmer, CQ and SQ were analyzed from the vowel sounds. For the voice training, sound fundamental frequency was displayed during the vocalization and singing a song using pitches analyzed from the EGG. The system implemented in this study could be used for vocal exercise.

• Korean Journal of Remote Sensing
• /
• v.22 no.2
• /
• pp.131-139
• /
• 2006

IKONOS images have the perspective geometry in CCD sensor line like aerial images with central perspective geometry. So the occlusion by buildings, terrain or other objects exist in the image. It is difficult to detect the occlusion with RPCs(rational polynomial coefficients) for ortho-rectification of image. Therefore, in this study, we detected the occlusion areas in IKONOS images using the nominal collection elevation/azimuth angle and restored the hidden areas using another stereo images, from which the rue ortho image could be produced. The algorithm's validity was evaluated using the geometric accuracy of the generated ortho image.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.1
• /
• pp.51.2-51.2
• /
• 2019

Multi-wavelength observations become very popular in astronomy. Even though there are some correlations among different sensor images, it is not easy to translate from one to the other one. In this study, we apply a deep learning method for image-to-image translation, based on conditional generative adversarial networks (cGANs), to solar images. To examine the validity of the method for scientific data, we consider several different types of pairs: (1) Generation of SDO/EUV images from SDO/HMI magnetograms, (2) Generation of backside magnetograms from STEREO/EUVI images, (3) Generation of EUV & X-ray images from Carrington sunspot drawing, and (4) Generation of solar magnetograms from Ca II images. It is very impressive that AI-generated ones are quite consistent with actual ones. In addition, we apply the convolution neural network to the forecast of solar flares and find that our method is better than the conventional method. Our study also shows that the forecast of solar proton flux profiles using Long and Short Term Memory method is better than the autoregressive method. We will discuss several applications of these methodologies for scientific research.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.33 no.6
• /
• pp.2551-2562
• /
• 2013

Since 2006, an Intelligent Excavation Robot which automatically performs the earth-work without operator has been developed in Korea. The technologies for automatically recognizing the terrain of work environment and detecting the objects such as obstacles or dump trucks are essential for its work quality and safety. In several countries, terrestrial 3D laser scanner and stereo vision camera have been used to model the local area around workspace of the automated construction equipment. However, these attempts have some problems that require high cost to make the sensor system or long processing time to eliminate the noise from 3D model outcome. The objectives of this study are to analyze the advantages of the existing 3D modeling sensors and to examine the applicability for practical use by using Analytic Hierarchical Process(AHP). In this study, 3D modeling quality and accuracy of modeling sensors were tested at the real earth-work environment.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.46 no.2
• /
• pp.61-71
• /
• 2009

In this paper, we propose a low-cost and compact motion capture system which enables to play motion games in PS2(Play Station 2). Recently, motion capture systems which are being used as a part in film producing and making games are too expensive and enormous systems. Now days, motion games using common USB camera are slow and have two-dimension recognition. But PSD sensor has a few good points, such as fast and low-cost. In recently year, 3D motion capture systems using 2D PSD (Position Sensitive Detector) optic sensor for motion capturing have been developed. One is Multi-PSD motion capture system applying stereo vision and another is Single-PSD motion capture system applying optical theory ship. But there are some problems to apply them to motion games. The Multi-PSD is high-cost and complicated because of using two more PSD Camera. It is so difficult to make markers having omni-direction equal intensity in Single-PSD. In this research, we propose a new theory that solves aforementioned problems. It can measure 3D coordination if separated two marker's intensity is equal to. We made a system based on this theory and experimented for performance capability. As a result, we were able to develop a motion capture system which is a single, low-cost, fast, compact, wide-angle and an adaptable motion games. The developed system is expected to be useful in animation, movies and games.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.46 no.3
• /
• pp.219-229
• /
• 2018

This paper proposes Kalman Filter-based relative navigation algorithms for proximity tasks such as rendezvous/docking/cluster-operation of spacecraft using PSD Sensors and Infrared Beacon Modules. Numerical simulations are performed for comparative analysis of the performance of each relative-navigation technique. Based on the operation principle and optical modeling of the PSD Sensor and the Infrared Beacon Module used in the relative navigation algorithm, a measurement model for the Kalman filter is constructed. The Extended Kalman Filter(EKF) and the Unscented Kalman Filter(UKF) are used as probabilistic relative navigation based on measurement fusion to utilize kinematics and dynamics information on translational and rotation motions of satellites. Relative position and relative attitude estimation performance of two filters is compared. Especially, through the simulation of various scenarios, performance changes are also investigated depending on the number of PSD Sensors and IR Beacons in target and chaser satellites.

• Journal of IKEEE
• /
• v.18 no.4
• /
• pp.636-644
• /
• 2014

This paper is to technically implement the sensing platform for Home-Wellness Robot. First, self-localization technique is based on a smart home and object in a home environment, and IOT(Internet of Thing) between Home Wellness Robots. RF tag is set in a smart home and the absolute coordinate information is acquired by a object included RF reader. Then bluetooth communication between object and home wellness robot provides the absolute coordinate information to home wellness robot. After that, the relative coordinate of home wellness robot is found and self-localization through a stereo camera in a home wellness robot. Second, this paper proposed fuzzy control methode based on a vision sensor for approach object of home wellness robot. Based on a stereo camera equipped with face of home wellness robot, depth information to the object is extracted. Then figure out the angle difference between the object and home wellness robot by calculating a warped angle based on the center of the image. The obtained information is written Look-Up table and makes the attitude control for approaching object. Through the experimental with home wellness robot and the smart home environment, confirm performance about the proposed self-localization and posture control method respectively.

• Journal of the Korea Computer Graphics Society
• /
• v.27 no.5
• /
• pp.63-71
• /
• 2021

Among drone autonomous flight technologies, obstacle avoidance is a very important technology that can prevent damage to drones or surrounding environments and prevent danger. Although the LiDAR sensor-based obstacle avoidance method shows relatively high accuracy and is widely used in recent studies, it has disadvantages of high unit price and limited processing capacity for visual information. Therefore, this paper proposes an obstacle avoidance algorithm for drones using camera-based PPO(Proximal Policy Optimization) reinforcement learning, which is relatively inexpensive and highly scalable using visual information. Drone, obstacles, target points, etc. are randomly located in a learning environment in the three-dimensional space, stereo images are obtained using a Unity camera, and then YOLov4Tiny object detection is performed. Next, the distance between the drone and the detected object is measured through triangulation of the stereo camera. Based on this distance, the presence or absence of obstacles is determined. Penalties are set if they are obstacles and rewards are given if they are target points. The experimennt of this method shows that a camera-based obstacle avoidance algorithm can be a sufficiently similar level of accuracy and average target point arrival time compared to a LiDAR-based obstacle avoidance algorithm, so it is highly likely to be used.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2004.04a
• /
• pp.337-343
• /
• 2004

Spatial information could be obtained from spaceborne high resolution optical and synthetic aperture radar(SAR) images. However, some satellite images do not provide physical sensor information instead, rational polynomial coefficients(RPC) are available. The objectives of this study are: (1) 3-dimensional ground coordinates were computed by applying rational function model(RFM) with the RPC for the stereo pair of Ikonos images and their accuracy was evaluated. (2) Interferometric SAR(InSAR) was applied to JERS-1 images to generate DEM and its accuracy was analysis. (3) Quality of the DEM generated automatically also analyzed for different types of terrain in the study site. The overall accuracy was evaluated by comparing with GPS surveying data. The height offset in the RPC was corrected by estimating bias. In consequence, the accuracy was improved. Accuracy of the DEMs generated from InSAR with different selection of GCP was analyzed. In case of the Ikonos images, the results show that the overall RMSE was 0.23327", 0.l1625" and 13.70m in latitude, longitude and height, respectively. The height accuracy was improved after correcting the height offset in the RPC. i.e., RMSE of the height was 1.02m. As for the SAR image, RMSE of the height was 10.50m with optimal selection of GCP. For the different terrain types, the RMSE of the height for urban, forest and flat area was 23.65m, 8.54m, 0.99m, respectively for Ikonos image while the corresponding RMSE was 13.82m, 18.34m, 10.88m, respectively lot SAR image.