• Proceedings of the KSRS Conference
• /
• 2009.03a
• /
• pp.147-150
• /
• 2009

본 논문에서는 해상 클러터를 고려하여 움직이는 물체의 SAR 원시 데이터를 생성하고, SAR 원시 데이터 중간 처리 결과인 range 압축 데이터의 azimuth 차분 신호로부터 물체의 속도를 측정하는 방법을 여러 가지 환경에 적용하여 그 정확도 및 적용 가능한 경우를 분석하였다. 움직이는 물체에 의한 도플러 중심 주파수의 변이가 azimuth 차분 신호에서 위상의 변화를 가져오므로, 이를 이용하여 움직이는 물체의 속도를 측정하는 알고리듬을 정리하였다. 이 알고리듬을 위에서 생성한 range 압축 데이터에 적용하여, 타깃이 되는 물체가 독립적으로 존재하는 경우, azimuth 상에 또 다른 속도를 가지는 산란체가 존재하는 경우, 그리고 높은 후방산란계수를 가지는 육지에 타깃이 되는 물체가 인접해 있는 경우를 가정하여 속도를 측정하였다. 그 결과, 타깃이 되는 물체가 SAR 영상에서 256 픽셀 범위 내에서 독립적으로 존재할 경우에는 높은 정확도로 물체의 속도를 측정할 수 있었으나, 128 픽셀 범위에 다른 움직이는 물체가 존재하거나, 높은 후방산란 계수를 갖는 육지와 인접해 있을 경우에는 최대 1m/s 의 오차를 나타냈다. 이는 주변 산란체의 영향에 의해 신호가 교란되어 목표물의 위치를 추정하는 과정에서 오차가 발생했기 때문이다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2009.10a
• /
• pp.565-568
• /
• 2009

This paper proposes an monitoring system of elder and feeble person's motion activity using an object's motion activity data. The proposed system used wireless 3-axis sensor module, product by Freescale(Wireless Sensing Triple Axis Reference Design Board (ZSTAR)). We distribute sensing data into three classes using Neural Network System SVM. We find performance of proposed system that simulate some case about walk, past walk, fallen. Classify result data and graph of sensing data present succes rate 80%.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.237-240
• /
• 2002

지면에 가까이 있는 비행체나 수면에 가까이 있는 수중 운동체에서 도래하는 신호는 직접파와 반사파로 구성되어 있다. 그런 환경에서는 반사파가 직접파와 coherent한 관계를 갖는다. 이런 현상으로 해서 DOA추정을 위해 추정한 correlation 행렬이 singular한 성질을 갖게 된다. 이렇게 singular한 행렬은 DOA추정에 큰 오류를 내게 된다. 또 움직이는 대상의 경우 추정 correlation 행렬의 성질이 시간에 따라 변하게 된다. 본 논문에서는 위 두 상황을 함께 해결하기 위해서 PASTd알고리즘을 변형하여 Forward/Backward Variable forgetting factor를 도입한 PASTd알고리즘을 제안한다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.4
• /
• pp.1294-1300
• /
• 2010

In an environment where all nodes move, the sensor node receives anchor node's position information within communication radius and modifies the received anchor node's position information by one's traveled distance and direction in saving in one's memory, where if there at least 3, one's position is determined by performing localization through trilateration. The proposed localization mechanisms have been simulated in the Matlab. In an environment where certain distance is maintained and nodes move towards the same direction, the probability for the sensor node to meet at least 3 anchor nodes with absolute coordinates within 1 hub range is remote. Even if the sensor node has estimated its position with at least 3 beacon information, the angle ${\theta}$ error of accelerator and digital compass will continuously apply by the passage of time in enlarging the error tolerance and its estimated position not being relied. Dead reckoning technology is used as a supplementary position tracking navigation technology in places where GPS doesn't operate, where one's position can be estimated by knowing the distance and direction the node has traveled with acceleration sensor and digital compass. The localization algorithm to be explained is a localization technique that uses Dead reckoning where all nodes are loaded with omnidirectional antenna, and assumes that one's traveling distance and direction can be known with accelerator and digital compass. The simulation results show that our scheme performed better than other mechanisms (e.g. MCL, DV-distance).

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.10C
• /
• pp.958-966
• /
• 2003

Change detection algorithms based on frame difference are frequently used for finding moving objects in image sequences. These algorithms detect the change of frames using estimated statistical background model. But, if this estimated background model is different from the actual statistical distribution, false detections are generated. In this paper, we propose an improved change detection algorithm using euler number and structure tensor. The proposed mapping method which is based on the euler number can be used for reducing the false detections that generated by nonparametric change detection algorithm. In this paper, the change in the region of moving object also can be detected by the proposed method using structure tensor. Experimental result shows that the proposed method reduces the false detections effectively by 90% on "Weather", by 34% on "Mother & daughter" and by 43% on "Aisle" than an existing method does.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.10
• /
• pp.173-179
• /
• 2014

In this paper, a motor parameter identification system has been implemented to improve the performance of the position sensorless stroke controller for linear compressors. In order to control the cooling capability of a refrigerator or an air conditioner in which linear compressors are applied, the piston speed should be controlled. The piston speed control can be obtained by adjusting the frequency or the stroke of linear motors. The dynamic performance of linear compressors depends on how accurately the stroke or the piston amplitude is estimated. The merits and demerits of Constant method and PIM (Parameter Identification Method) concerning the needed memory space and the stroke error are discussed and verified via some experimental studies.

• Journal of KIISE:Software and Applications
• /
• v.31 no.2
• /
• pp.173-181
• /
• 2004

In this paper, we propose a system for automatic moving object detection and tracking in sequence images acquired from a moving camera. The proposed algorithm consists of moving object detection and its tracking. Moving object can be detected by integration of BBME and DD method We segment the detected object using histogram back projection, match it using histogram intersection, extract and track it using XY-projection. Computer simulation results have shown that the proposed algorithm is reliable and can successfully detect and track a moving object on image sequences obtained by a moving camera.

• Proceedings of the Korea Contents Association Conference
• /
• 2005.11a
• /
• pp.485-489
• /
• 2005

본 연구에서는 작업자의 일정한 동작을 보다 효율적으로 인식할 수 있는 시스템을 제안하고자 한다. 먼저, 작업자의 동작을 촬영한 동영상에서 연속된 프레임간의 차를 기반으로, 고정된 배경과 움직이는 대상을 분리한다. 다음으로, 에지 검출을 이용하여 동작의 중심 위치를 추정하여 연속적으로 움직이는 동작을 인식할 수 있도록 하였다. 본 연구에서 설계한 동작 인식시스템은 기존의 산업현장에서 적용되고 있는 동작인식 시스템의 문제점을 보완하기 위하여 작업자의 동작을 고정된 CCTV 로 촬영한 영상을 인식의 대상으로 취함으로써 동작 정보를 얻기 위한 각종 장비들이 최소화되었다. 또한, 작업자의 신체 부분별 특성을 추출하기 위한 계산작업에 소요되는 시간을 줄이기 위하여 프레임간의 차연산과 에지검출을 통한 동작인식을 실시하여 인식에 필요한 작업시간을 단축하여, 효율적이면서 비용이 저렴한 동작 인식시스템을 설계하였다.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2001.06a
• /
• pp.217-220
• /
• 2001

In this paper, we propose a segmentation method of moving object using the spatio-temporal information in image sequences. Proposed method consists of motion detection step using difference image, region segmentation step using k-means algorithm, motion estimation step and segmenting step using intensity and motion information. Experimental results show that the method is capable of segmenting variously moving objects in image sequences.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.3
• /
• pp.17-24
• /
• 2004

In this paper, the wavelet video coding using Low-Band-Shift(LBS) method and multiresolution motion estimation(MRME) is proposed. To overcome shift- variant property on wavelet coefficients, the LBS was proposed. LBS method previously has superior performance in terms of rate-distortion characteristic. However, this method needs more memory and computational complexity. Therefore to reduce computational complexity of video coding using LBS, we combine MRME with LBS. When mm is applied only, it has 7 times as much as existing method's motion vector because each subband has different motion vector using property of LBS, number of motion vector decreases. Proposed method decreases motion vector, and it decreases motion compensated Prediction error by detailed motion estimation. And then it shows better coding performance. Also this method reduces computational amount by smaller search area in higher resolution. The computational complexity of the proposed method is 12.1% of that of existing method at 3-level wavelet transform. The experimental results with the proposed method show about 0.2∼9.7% improvement of MAD performance in case of lossless coding, and 0.1∼2.0㏈ improvement of PSNR performance at 4he same bit rate in case of lossy coding.