• Journal of Korea Society of Industrial Information Systems
• /
• v.16 no.2
• /
• pp.65-72
• /
• 2011

In this paper, a real-time monocular vision based navigation system is developed for the disabled people, where online background learning and vector field histogram are used for identifying obstacles and recognizing avoidable paths. The proposed system is performed by three steps: obstacle classification, occupancy grid map generation and VFH-based path recommendation. Firstly, the obstacles are discriminated from images by subtracting with background model which is learned in real time. Thereafter, based on the classification results, an occupancy map sized at $32{\times}24$ is produced, each cell of which represents its own risk by 10 gray levels. Finally, the polar histogram is drawn from the occupancy map, then the sectors corresponding to the valley are chosen as safe paths. To assess the effectiveness of the proposed system, it was tested with a variety of obstacles at indoors and outdoors, then it showed the a'ccuracy of 88%. Moreover, it showed the superior performance when comparing with sensor based navigation systems, which proved the feasibility of the proposed system in using assistive devices of disabled people.

• The KIPS Transactions:PartB
• /
• v.16B no.1
• /
• pp.35-46
• /
• 2009

Lane extraction and lane departure warning algorithms using the image sensor attached in the vehicle are addressed. With the research about intelligent automobile, there have been many algorithms about lane recognition and lane departure warning system. However, since these algorithms require to detect 2 lanes, the high time complexity and the low recognition rate under various driving circumstances are critical problems. In this paper, we present a lane departure warning algorithm using single lane extraction and center point analysis that achieves the fast processing time and high detection rate. From the geometry between camera and objects, the region of interest (ROI) is determined and splitted into two parts. Hough transform detects the part of the lane. After the detected lane is restored to have a pre-determined size, lane departure is estimated by calculating the distance from the center point. On real driving environments, the presented algorithm is compared with previous algorithms. Experiment results support that the presented algorithm is fast and accurate.

• Proceedings of the Korean Information Science Society Conference
• /
• 2010.06a
• /
• pp.256-257
• /
• 2010

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.06a
• /
• pp.5-8
• /
• 2019

단일 영상 초해상도 (Single Image Super-Resolution - SISR)기법은 카메라로 획득된 저해상도 영상에 필터 기반의 연산을 적용하여 좋은 화질의 고해상도 영상을 복원하는 과정이다. 최근에 심층 합성곱 신경망 학습의 발전에 따라 단일 영상 초해상도에 적용되는 심층 학습 기법들은 좋은 성과를 보여 주고 있다. 본 논문은 단일 영상 초해상도 성능을 개선하기 위해 웨이블릿 예측 네트워크를 효율적으로 적용하는 방법에 대해 연구하였으며, 저해상도 입력 영상의 특징을 잘 추출해내기 위해 네트워크 내부에 RDB를 적용하여 기존 방식보다 효율적으로 고해상도 영상 복원하는 기법을 제안한다. 모의실험을 통해 제안하는 방법이 기존 방법보다 화질은 약 PSNR 0.18dB만큼 우수하며 속도는 1.17배 빠른 것을 확인하였다.

• Journal of KIISE:Computing Practices and Letters
• /
• v.13 no.7
• /
• pp.488-498
• /
• 2007

This paper presents a novel approach for representing the spatio-temporal topology of the camera network with overlapping and non-overlapping fields of view (FOVs) in Ubiquitous Smart Space (USS). The topology is determined by tracking moving objects and establishing object correspondence across multiple cameras. To track people successfully in multiple camera views, we used the Merge-Split (MS) approach for object occlusion in a single camera and the grid-based approach for extracting the accurate object feature. In addition, we considered the appearance of people and the transition time between entry and exit zones for tracking objects across blind regions of multiple cameras with non-overlapping FOVs. The main contribution of this paper is to estimate transition times between various entry and exit zones, and to graphically represent the camera topology as an undirected weighted graph using the transition probabilities.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.41 no.5
• /
• pp.422-428
• /
• 2013

Recently, many unmanned aerial vehicle (UAV) studies have focused on small UAVs, because they are cost effective and suitable in dangerous indoor environments where human entry is limited. Map building through distance measurement is a key technology for the autonomous flight of small UAVs. In many researches for unmanned systems, distance could be measured by using laser range finders or stereo vision sensors. Even though a laser range finder provides accurate distance measurements, it has a disadvantage of high cost. Calculating the distance using a stereo vision sensor is straightforward. However, the sensor is large and heavy, which is not suitable for small UAVs with limited payload. This paper suggests a low-cost distance measurement system using a laser pointer and a monocular vision sensor. A method to measure distance using the suggested system is explained and some experiments on map building are conducted with these distance measurements. The experimental results are compared to the actual data and the reliability of the suggested system is verified.

• Journal of KIISE:Software and Applications
• /
• v.35 no.12
• /
• pp.780-788
• /
• 2008

Modeling hand poses and tracking its movement are one of the challenging problems in computer vision. There are two typical approaches for the reconstruction of hand poses in 3D, depending on the number of cameras from which images are captured. One is to capture images from multiple cameras or a stereo camera. The other is to capture images from a single camera. The former approach is relatively limited, because of the environmental constraints for setting up multiple cameras. In this paper we propose a method of reconstructing 3D hand poses from a 2D input image sequence captured from a single camera by means of Belief Propagation in a graphical model and recognizing a finger clicking motion using a hidden Markov model. We define a graphical model with hidden nodes representing joints of a hand, and observable nodes with the features extracted from a 2D input image sequence. To track hand poses in 3D, we use a Belief Propagation algorithm, which provides a robust and unified framework for inference in a graphical model. From the estimated 3D hand pose we extract the information for each finger's motion, which is then fed into a hidden Markov model. To recognize natural finger actions, we consider the movements of all the fingers to recognize a single finger's action. We applied the proposed method to a virtual keypad system and the result showed a high recognition rate of 94.66% with 300 test data.

• Proceedings of the Korea Society for Industrial Systems Conference
• /
• 1999.12a
• /
• pp.493-505
• /
• 1999

Zoom lens camera calibration is an important and difficult problem for two reason at least. First, the intrinsic parameters of such a camera change over time, it is difficult to calibration them on-line. Secondly, the pin-hole model for single lens system can not be applied directly to a zoom lens system. In this paper, We address some aspects of this problem.

• Proceedings of the IEEK Conference
• /
• 2009.05a
• /
• pp.207-209
• /
• 2009

다양한 센서 및 영상 카메라를 이용한 교통, 보안 및 안전 감시 시스템에 있어 처리해야 하는 영상 데이터의 양은 점점 커져가고 있다. 또한 단일 카메라가 아닌 많은 수의 카메라를 이용할 경우 운영자가 모든 영상 데이터를 확인하고 이에 대한 응답을 즉시 하기가 힘이 든다. 따라서 영상 데이터를 이용하기 위한 시스템에서 소프트웨어적인 처리는 필수이며 물체를 정확하게 추적하기 위해서는 물체를 인식하고 물체의 움직임을 예측하고 움직임을 보정하는 단계가 필요하다. 본 논문에서는 물체의 움직임을 정확히 추적하기 위하여 이동 물체를 추적할 때에 적절한 Kalman 필터를 이용하여 고속 물체 추적 시스템을 구현하였다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.10a
• /
• pp.741-744
• /
• 2003

in this paper, we propose a dynamic-based compression system by creating mosaic background and transmitting the change information. A dynamic mosaic of the background is progressively integrated in a single image using the camera motion information. For the camera motion estimation, we calculate perspective projection parameters for each frame sequentially with respect to its previous frame. The camera motion is robustly estimated on the background by discriminating between background and foreground regions. The modified block-based motion estimation is used to separate the background region.