• Journal of the Institute of Electronics and Information Engineers
• /
• v.49 no.9
• /
• pp.300-306
• /
• 2012

In this paper, we propose a vision based hand motion recognition method for an untouchable user interface of smart devices. First, an original color image is converted into a gray scaled image and its spacial resolution is reduced, taking the small memory and low computational power of smart devices into consideration. For robust recognition of hand motions through separation of horizontal and vertical motions, the horizontal principal area (HPA) and the vertical principal area (VPA) are defined respectively. From the difference images of the consecutively obtained images, the center of gravity (CoG) of the significantly changed pixels caused by hand motions is obtained, and the direction of hand motion is detected by defining the least mean squared line for the CoG in time. For verifying the feasibility of the proposed method, the experiments are carried out with a vision system.

• Health Policy and Management
• /
• v.16 no.3
• /
• pp.52-69
• /
• 2006

Researchers indicates several issues as to performance evaluation methods for public health centers. Firstly, health centers are passively involved in the evaluation process, the performance indicators are activity-oriented, and mostly unrealistic. Balanced Scorecard is one of the methods for evaluating organizational performance, which is utilized at current in many industries including public sector. The purposes of this study is to apply balanced scorecard approach to a public health center and to develop performance indicators as well as their vision and strategies. For developing performance indicators, researchers selected K City Public Health Center and implemented brainstorming with members of health promotion team. Through team process they suggested goals, strategies and several indicators they considered proper to accomplish their shared vision. And then appropriateness of the indicators were evaluated by professional researchers in health promotion field for consensus building by email questionnaire. Based on survey and professional consensus meeting, 11 performance indicators were tailored in four perspectives as well as 6 strategies and 10 strategic goals, which are steps for accomplishing shared vision of health promotion team. For details, refer to the paper. Most members of health promotion team were satisfied with the indicators. However issues such as low level of recognition and familiarity with a new concept of BSC, and cultural resistance to strategic approach in public organizations should be addressed for future application and dissemination of BSC technique to public health organizations.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.38 no.4
• /
• pp.342-351
• /
• 2010

UAV (Unmanned Aerial Vehicle) have an INS(Inertial Navigation System) equipment and also have an electro-optical Equipment for mission. This paper proposes the vision based attitude estimation algorithm using Kalman Filter and Optical flow for UAV. Optical flow is acquired from the movie of camera which is equipped on UAV and UAV's attitude is measured from optical flow. In this paper, Kalman Filter has been used for the settlement of the low reliability and estimation of UAV's attitude. Algorithm verification was performed through experiments. The experiment has been used rate table and real flight video. Then, this paper shows the verification result of UAV's attitude estimation algorithm. When the rate table was tested, the error was in 2 degree and the tendency was similar with AHRS measurement states. However, on the experiment of real flight movie, maximum yaw error was 21 degree and Maximum pitch error was 7.8 degree.

• The Journal of Korea Robotics Society
• /
• v.6 no.3
• /
• pp.197-209
• /
• 2011

This paper presents a localization system using ceiling images in a large indoor environment. For a system with low cost and complexity, we propose a single camera based system that utilizes ceiling images acquired from a camera installed to point upwards. For reliable operation, we propose a method using hybrid features which include natural landmarks in a natural scene and artificial landmarks observable in an infrared ray domain. Compared with previous works utilizing only infrared based features, our method reduces the required number of artificial features as we exploit both natural and artificial features. In addition, compared with previous works using only natural scene, our method has an advantage in the convergence speed and robustness as an observation of an artificial feature provides a crucial clue for robot pose estimation. In an experiment with challenging situations in a real environment, our method was performed impressively in terms of the robustness and accuracy. To our knowledge, our method is the first ceiling vision based localization method using features from both visible and infrared rays domains. Our system can be easily utilized with a variety of service robot applications in a large indoor environment.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.6
• /
• pp.621-628
• /
• 2015

This paper describes an algorithm for detecting vehicles detection in real time. The proposed algorithm has the technique based on computer vision and image processing. In real, complex environment such as one with road traffic, many algorithms have great difficulty such as low detection rate and increasing computational time due to complex backgrounds and rapid changes. To overcome this problem in this paper, the proposed algorithm consists of the following methods. First, to effectively separate the candidate regions, we use vertical and horizontal edge information, and shadow values from input image sequences. Second, we extracts features by using HOG from the selected candidate regions. Finally, this paper uses the OS fuzzy-ELM based on SLFN to classify the extracted features. The experimental results show that the proposed method perform well for detecting vehicles and improves the accuracy and the computational time of detecting.

• Journal of the Microelectronics and Packaging Society
• /
• v.23 no.3
• /
• pp.37-41
• /
• 2016

In this paper, we introduced a near infrared fluorescence imaging system that has long working distance and analyzed on the effects of measurement variables such as gain, exposure time, working distance, magnification. Fluorescence signal intensity is growing up according to exposure time and magnification increasing, and it is getting stronger according to increase of gain, but the background signal intensity is getting stronger together. It causes low SBR. Due to a laser irradiation method, laser intensity distribution of the introduced system is not uniform and it makes fluorescence signal weak. So, we proposed a solution.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.9
• /
• pp.281-290
• /
• 2020

The Vision-and-Language Navigation(VLN) task is a complex intelligence problem that requires both visual and language comprehension skills. In this paper, we propose a new learning model for visual-language navigation agents. The model adopts a hybrid learning that combines imitation learning based on demo data and reinforcement learning based on action reward. Therefore, this model can meet both problems of imitation learning that can be biased to the demo data and reinforcement learning with relatively low data efficiency. In addition, the proposed model uses a novel path-based reward function designed to solve the problem of existing goal-based reward functions. In this paper, we demonstrate the high performance of the proposed model through various experiments using both Matterport3D simulation environment and R2R benchmark dataset.

• Korean Institute of Interior Design Journal
• /
• v.18 no.1
• /
• pp.35-42
• /
• 2009

In this study, the author aims to examine that the image for specific space coming through vision is to reveal how people perceive the space through vision and whether the perceived space includes the role as the catalyst that causes the following activities. It is believed that the fact which elements are remembered as the whole expression as well as the memorized images that humans have regarding the space should act as the important factor in terms of space perception. The conclusions from this study are as follow; 1) By analyzing the frequency of observation time that were obtained from the examinee, it was possible to classify the property of observation with S areas. Besides, it was possible to the meanings that the design elements have in each area. The establishment of the areas are considered as the important factor to examine which design elements have drawn the attention. 2) In case of I area which showed the most design factors that would lead examinee's vision or have interests in the examinee views, it showed that it stared the lower parts from the middle of the Image spatially, which was the most stable position from the image with strong tendency for staring at this area. 3) The most frequently stared area was the lower part of the middle, however, while the I area gazed the right side of the middle, II area faced the left side more so that it was revealed that it stared at the lower part of the middle and right side, then, moved to left. 4) Despite the frequent observation, some areas had very low or few observation data records and the area which was designed with monotonous color with relatively large size was also involved here so that it was identified that the simply treated area in design was rarely gazed.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.8
• /
• pp.113-119
• /
• 2013

The requirement of rehabilitation is increasing from the stroke, spinal cord injury. One of the most difficult part is the upper limb rehabilitation because of its nervous complexity. A rehabilitation has effectiveness when a professional therapist treats in work at facility, but it has problems of an accessibility, a constant availability, a self-participation and taking lots of cost and time. In this paper, we test and experiment the accuracy and execution time of the pattern recognition algorithms like PCA, ICA, LDA, SVM to show the recognition possibility of it on the upper limb function from the 3D trajectory information which is gathered from stereo vision sensor(Kinect). From the result, PCA, ICA have low accuracy, but LDA, SVM have good accuracy to use for physical rehabilitation on the upper limb function.

• Smart Structures and Systems
• /
• v.31 no.4
• /
• pp.325-334
• /
• 2023

Computer vision-based damage detection enables non-contact, efficient and low-cost bridge health monitoring, which reduces the need for labor-intensive manual inspection or that for a large number of on-site sensing instruments. By leveraging recent semantic segmentation approaches, we can detect regions of critical structural components and identify damages at pixel level on images. However, existing methods perform poorly when detecting small and thin damages (e.g., cracks); the problem is exacerbated by imbalanced samples. To this end, we incorporate domain knowledge to introduce a hierarchical semantic segmentation framework that imposes a hierarchical semantic relationship between component categories and damage types. For instance, certain types of concrete cracks are only present on bridge columns, and therefore the noncolumn region may be masked out when detecting such damages. In this way, the damage detection model focuses on extracting features from relevant structural components and avoid those from irrelevant regions. We also utilize multi-scale augmentation to preserve contextual information of each image, without losing the ability to handle small and/or thin damages. In addition, our framework employs an importance sampling, where images with rare components are sampled more often, to address sample imbalance. We evaluated our framework on a public synthetic dataset that consists of 2,000 railway bridges. Our framework achieves a 0.836 mean intersection over union (IoU) for structural component segmentation and a 0.483 mean IoU for damage segmentation. Our results have in total 5% and 18% improvements for the structural component segmentation and damage segmentation tasks, respectively, compared to the best-performing baseline model.