• Journal of KIISE:Software and Applications
• /
• v.30 no.12
• /
• pp.1228-1238
• /
• 2003

This paper presents a new and practical method based on computer vision for detecting the monitor position where the user is looking. In general, the user tends to move both his face and eyes in order to gaze at certain monitor position. Previous researches use only one wide view camera, which can capture a whole user's face. In such a case, the image resolution is too low and the fine movements of user's eye cannot be exactly detected. So, we implement the gaze detection system with dual camera systems(a wide and a narrow view camera). In order to locate the user's eye position accurately, the narrow view camera has the functionalities of auto focusing and auto panning/tilting based on the detected 3D facial feature positions from the wide view camera. In addition, we use dual R-LED illuminators in order to detect facial features and especially eye features. As experimental results, we can implement the real-time gaze detection system and the gaze position accuracy between the computed positions and the real ones is about 3.44 cm of RMS error.

• Journal of Broadcast Engineering
• /
• v.26 no.5
• /
• pp.519-532
• /
• 2021

Recently, in various fields such as games, movies, and animation, content that uses motion capture to build body models and create characters to express in 3D space is increasing. Studies are underway to generate animations using RGB-D cameras to compensate for problems such as the cost of cinematography in how to place joints by attaching markers, but the problem of pose estimation accuracy or equipment cost still exists. Therefore, in this paper, we propose a system that inputs RGB images into a joint estimation network and converts the results into 3D data to create FBX format animations in order to reduce the equipment cost required for animation creation and increase joint estimation accuracy. First, the two-dimensional joint is estimated for the RGB image, and the three-dimensional coordinates of the joint are estimated using this value. The result is converted to a quaternion, rotated, and an animation in FBX format is created. To measure the accuracy of the proposed method, the system operation was verified by comparing the error between the animation generated based on the 3D position of the marker by attaching a marker to the body and the animation generated by the proposed system.

• The Journal of Korean Society for Radiation Therapy
• /
• v.29 no.1
• /
• pp.69-76
• /
• 2017

Purpose: The purpose of this study was to verify dosimetric results and reproducibility of position during craniospinal irradiation (CSI) using tomotherapy (Accuray Incorporated, USA). Also, by comparing with conventional CSI Technique, we confirmed the efficiency of using a Tomotherapy. Materials and Methods: 10 CSI patients who get tomotherapy participate. Patient-specific quality assurances (QA) for each patient are conducted before treatment. When treating, we took Megavoltage Computed Tomography (MVCT) that range of head and neck before treatment, L spine area after treatment. Also we conducted in-vivo dosimetry to check a scalp dose. Finally, we made a 3D conventional radiation therapy(3D-CRT) of those patients to compare dosimetric differences with tomotherapy treatment planning. Results: V107, V95 of brain is 0 %, 97.2 % in tomotherapy, and 0.3 %, 95.1 % in 3D-CRT. In spine, value of V107, V95 is 0.2 %, 18.6 % in tomotherapy and 89.6 %, 69.9 % in 3D-CRT. Except kidney and lung, tomotherapy reduced normal organ doses than 3D-CRT. The maximum positioning error value of X, Y, Z was 10.2 mm, -8.9 mm, -11.9 mm. Through in-vivo dosimetry, the average of scalp dose was 67.8 % of prescription dose. All patient-specific QA were passed by tolerance value. Conclusion: CSI using tomotherapy had a risk of parallel organ such as lung and kidney because of integral dose in low dose area. However, it demonstrated dosimetric superiority at a target and saved normal organ to reduce high dose. Also results of reproducibility were not exceeded margins that estimated treatment planning and invivo dosimetry showed to reduce scalp dose. Therefore, CSI using tomotherapy is considered to efficient method to make up for 3D-CRT.

• The Journal of Korean Society for Radiation Therapy
• /
• v.35
• /
• pp.7-13
• /
• 2023

Purpose: In this study, we evaluated the effect of using a customized bolus on dose delivery in the treatment plan when cervical cancer protruded out of the body along with the uterus and evaluated reproducibility in patient set-up. Materials & Methods: The treatment plan used the Eclipse Treatment Planning System (Version 15.5.0, Varian, USA) and the treatment machine was VitalBeam (Varian Medical Systems, USA). The radiotherapy technique used 6 MV energy in the AP/PA direction with 3D-CRT. The prescribed dose is 1.8 Gy/fx and the total dose is 50.4 Gy/28 fx. Semiflex TM31010 (PTW, Germany) was used as the ion chamber, and the dose distribution was analyzed and evaluated by comparing the planned and measured dose according to each position movement and the tumor center dose. The first measurement was performed at the center by applying a customized bolus to the phantom, and the measurement was performed while moving in the range of -2 cm to +2 cm in the X, Y, and Z directions from the center assuming a positional error. It was measured at intervals of 0.5 cm, the Y-axis direction was measured up to ±3 cm, and the situation in which Bolus was set-up incorrectly was also measured. The measured doses were compared based on doses corrected to CT Hounsfield Unit (HU) 240 of silicon instead of the phantom's air cavity. Result: The treatment dose distribution was uniform when the customized bolus was used, and there was no significant difference between the prescribed dose and the actual measured value even when positional errors occurred. It was confirmed that the existing sheet-type bolus is difficult to compensate for irregularly shaped tumors protruding outside the body, but customized Bolus is found to be useful in delivering treatment doses uniformly.

• The korean journal of orthodontics
• /
• v.29 no.5 s.76
• /
• pp.521-534
• /
• 1999

The purpose of this study was to find the difference of stress distribution on canine altered by the application point of preangulated T-loop spring. For this study, the finite element models of upper left canine, upper left second premolar and upper left first molar were made. Also, the finite element models of $0.017{\times}0.025$ inch preangulated, preactivated T-loop spring and $0.018{\times}0.025$ inch stainless steel wire were made. Three types of T-loop spring were made . the middle of activated T-loop is positioned in accordance with the middle position of distance of bracket position of both the canine and first molar, 2mm anterior, 2mm posterior. We compared the forces and the distribution of stress that were generated by the difference of position of T-loop spring. The results were as follows. 1. All of the 3 types of T-loop spring showed the similar retraction forces. 2. All showed the similar amount & pattern of stress distribution. 3. The centers of rotation of canine in 3 types of T-loop spring were same and were positioned between C and D plane. 4. The canine showed the intrusive force by 2mm anterior positioned T-loop spring, but the extrusive force by 2mm posterior positioned T-loop suing. Neverthless, because of the small amount of the forces, the effect of vertical force was not significant.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2005.11a
• /
• pp.190-200
• /
• 2005

In this paper, the image modelling of road's lane markings is established using view frustum(VF) modeling. This algorithm also involve the real time processing of the 3D position coordinate and the distance data from the camera to the points on the 3D world coordinate by the camera calibration. In order to reduce their measurement error, an useful algorithm for which analyze the geometric variations clue to traveling vehicle's fluctuation using VF model is proposed. In experiments, without correction, for instance, the $0.4^{\circ}$ of pitching rotation gives the error of $0.4^{\sim}0.6m$ at the distance of 10m, but the more far distance cause exponentially the more error. We confirmed that this algorithm can be reduced less than 0.1m of error at the same condition.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.6 no.10
• /
• pp.2679-2691
• /
• 2012

In this paper, a high accuracy stereo reconstruction method for surgery instruments positioning is proposed. Usually, the problem of surgical instruments reconstruction is considered as a basic task in computer vision to estimate the 3-D position of each marker on a surgery instrument from three pairs of image points. However, the existing methods considered the 3-D reconstruction of the points separately thus ignore the structure information. Meanwhile, the errors from light variation, imaging noise and quantization still affect the reconstruction accuracy. This paper proposes a method which takes the structure information of surgical instruments as constraints, and reconstructs the whole markers on one surgical instrument together. Firstly, we calibrate the instruments before navigation to get the structure parameters. The structure parameters consist of markers' number, distances between each markers and a linearity sign of each instrument. Then, the structure constraints are added to stereo reconstruction. Finally, weighted filter is used to reduce the jitter. Experiments conducted on surgery navigation system showed that our method not only improve accuracy effectively but also reduce the jitter of surgical instrument greatly.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.12
• /
• pp.1167-1172
• /
• 2015

This paper describes the implementation of wearable AHRS for an electromagnetic motion capture system that can trace and analyze human motion on the principal nine axes of inertial sensors. The module provides a three-dimensional (3D) attitude and heading angles combining MEMS gyroscopes, accelerometers, and magnetometers based on the extended Kalman filter, and transmits the motion data to the 3D simulation via Wi-Fi to realize the unrestrained movement in open spaces. In particular, the accelerometer in AHRS is supposed to measure only the acceleration of gravity, but when a sensor moves with an external linear acceleration, the estimated linear acceleration could compensate the accelerometer data in order to improve the precision of measuring gravity direction. In addition, when an AHRS is attached in an arbitrary position of the human body, the compensation of the axis of rotation could improve the accuracy of the motion capture system.

• Journal of the Korean Society for Heat Treatment
• /
• v.30 no.3
• /
• pp.113-116
• /
• 2017

100 nm thick Sn doped ZnO (ZTO) single layer, 15 nm thick Ag buffered ZTO (ZTO/Ag), Ag intermediated ZTO (ZTO/Ag/ZTO) and Ag capped ZTO (Ag/ZTO) films were prepared on poly-carbonate (PC) substrates by RF and DC magnetron sputtering and then the influence of the Ag thin film on the optical and electrical properties of ZTO films were investigated. As deposited ZTO thin films show the visible transmittance of 81.8%, while ZTO/Ag/ZTO trilayer films show a higher visible transmittance of 82.5% in this study. From the observed results, it can be concluded that the 15 nm thick Ag interlayer enhances the opto-electrical performance of ZTO thin films effectively for use as flexible transparent conducting oxides films in various opto-electrical applications.

• 제어로봇시스템학회:학술대회논문집
• /
• 1991.10b
• /
• pp.1724-1729
• /
• 1991

This paper gives an overview of HIGIPS design concepts and prototype HIGIPS configuration, and discusses its application to recognition of the 3-D motion of a human arm. HIGIPS which employs the combination of pipeline architecture and multiprocessor architecture, is a high-speed, high-performance and low cost N * M multimicroprocessor parallel machine, where N is the number of pipeline stages and M is the number of processors in each stage. The algorithm to recognize the motion of a human arm with a single TV camera was developed on personal computer (NEC PC9801 series). As a constraint condition, some simple ring marks are used. Each joint of the arm is attached with a ring mark to obtain its centroid position when the arm moves. These centroid positions in the three-dimensional space are linked at each of the successive pictures of the moving arm to recover its overall motion. This algorithm takes about 2 seconds to process one image frame on the general-purpose personal computer. This paper mainly discuses how to partition this algorithm and execute on HIGIPS, and shows the speed up. From this application, it is clear that HIGIPS is an efficient machine for image processing and recognizing.