• IEIE Transactions on Smart Processing and Computing
• /
• v.2 no.4
• /
• pp.189-196
• /
• 2013

This paper presents a two-stage approach to detect pedestrians in video sequences taken from a moving vehicle. The first stage is a preprocessing step, in which potential pedestrians are hypothesized. During the preprocessing step, a difference image is constructed using a global motion estimation, vertical and horizontal edge maps are extracted, and the color difference between the road and pedestrians are determined to create candidate regions where pedestrians may be present. The candidate regions are refined further using the vertical edge symmetry features of the pedestrians' legs. In the next stage, each hypothesis is verified using the integral channel features and an AdaBoost classifier. In this stage, a decision is made as to whether or not each candidate region contains a pedestrian. The proposed algorithm was tested on a range of dataset images and showed good performance.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.01a
• /
• pp.463-464
• /
• 2019

본 논문은 기계학습을 기반으로 아두이노와 시리얼통신을 통한 사용자의 동작인식을 이용해 보다 간단하게 인터넷상의 원하는 동영상을 찾을 수 있는 검색시스템을 제작하고자 하였다. 이 검색시스템은 Python을 기반으로 SVM(Support Vector Machine)을 이용한 패턴 분류를 사용하였으며 이를 통해 사용자의 동작을 입력받아 문자를 예측 할 수 있다. 사용자는 이 검색시스템을 사용하기 위하여 우선 문자에 대한 사용자의 동작입력을 통해 학습 데이터 셋을 만들어야 하며 그것을 SVM을 이용하여 학습 모델과 식별자를 만들고, 만들어진 분류기를 통하여 동작인식을 바탕으로 문자의 결과를 예측 할 수 있다. 최종적으로 사용자의 동작인식을 거쳐 만들어진 문자열을 이용해 인터넷 동영상 사이트인 Youtube를 통해 웹 크롤링하여 문자열과 관련 있는 동영상을 찾아준다.

• Journal of Advanced Navigation Technology
• /
• v.9 no.1
• /
• pp.28-33
• /
• 2005

In this paper, we developed the algorithm to control longitudinal and lateral motion of UAV(Unmanned Aerial Vehicle) with Infrared heat sensors and GPS(Global Positioning System) receiver. UAV was controlled to be flown horizontally and also turned coordinately maintaining the constant altitude. Accomplishing the flight test of UAV sevral times, we were able to develope low price controller to control bank angle for lateral motion, and also pitch angle and altitude for longitudinal motion simultaneously.

• Journal of Astronomy and Space Sciences
• /
• v.22 no.2
• /
• pp.89-112
• /
• 2005

We determine the rotation velocities of 108 spiral and irregular galaxies (XV-Sample) from first-order rotation curves from position-velocity maps, based on short 21-cm observations with the Westerbork Synthesis Radio Telescope (WSRT). To test the usual random motion corrections, we compare the global HI linewidths and the rotation velocities, obtained from kinematical fits to two-dimensional velocity fields for a sample of 28 galaxies (RC-Sample), and find that the most frequently used correction formulae (Tully & Fouque 1985) are not very satisfactory. The rotation velocity parameter (the random-motion corrected HI linewidth: W?), derived with these corrections, may be statistically equal to two times the true rotation velocity, but in individual cases the differences can be large. We analyse, for both RC- and XV-Samples, the dependence of the slope of, and scatter in the Tully-Fisher relation on the definition of the rotation velocity parameters- For the RC-Sample, we find that the scatter in the Tully-Fisher relation can be reduced considerably when the rotation velocities derived from rotation curves are used instead of the random-motion corrected global H I linewidths. No such reduction in the scatter is seen for XV-Sample. We conclude that the reduction of the scatter in the Tully-Fisher relation seems to be related to the use of two-dimensional velocity information: accurate rotation velocity and kinematical inclination.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.1054-1059
• /
• 2005

The artificial potential method uses a potential field to guide a robot from a start to a goal configuration respectively. The potential field consists of attractive potential used to pull a robot toward a goal and repulsive potential to keep it away from obstacles. However, there are two problems concerning local minimum and computational cost to be resolved in conventional artificial potential methods. This study proposes a method utilizing a spline surface that interpolates arbitrary boundaries and a domain reduction method that reduces the unnecessary area. The proposed spline surface interpolates arbitrary shaped boundaries and is used as an artificial potential to guide a robot for global motion planning of a mobile robot. A reduced domain process reduces the unnecessary domain. We apply a distance-weighted function as such a function, which blends distances from each boundary with a reduction in computational time compared with other analytical methods. As a result, this paper shows that an arbitrary boundary spline surface provides global planning and a domain reduction method reduces local minimum with quick operation.

• Journal of Korea Multimedia Society
• /
• v.6 no.7
• /
• pp.1121-1130
• /
• 2003

This paper proposes a new motion estimation algorithm using the temporal continuity of motion. We set up a wide global search region (GSR), which basically corresponds to the search region of FSA, and local search regions (LSRs), the positions for which are predicted by the motion vectors of the temporal neighbor blocks, are constructed in the GSR. The previous frame blocks that possibly have effects on the current block are to be the temporal neighbor blocks. Then Motion estimation is only performed in the areas made by LSRs. Experimental results show that the proposed method can maintain visual qualifies with significant reductions of complexity by reducing search regions, when compared to the conventional FSA.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.9
• /
• pp.4438-4460
• /
• 2017

The Kernelized Correlation Filters (KCF) tracker has caused the extensive concern in recent years because of the high efficiency. Numerous improvements have been made successively. However, due to the abrupt motion between the consecutive image frames, these methods cannot track object well. To cope with the problem, we propose an extended KCF tracker based on swarm intelligence method. Unlike existing KCF-based trackers, we firstly introduce a swarm-based sampling method to KCF tracker and design a unified framework to track smooth or abrupt motion simultaneously. Secondly, we propose a global motion estimation method, where the exploration factor is constructed to search the whole state space so as to adapt abrupt motion. Finally, we give an adaptive threshold in light of confidence map, which ensures the accuracy of the motion estimation strategy. Extensive experimental results in both quantitative and qualitative measures demonstrate the effectiveness of our proposed method in tracking abrupt motion.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.9
• /
• pp.4412-4428
• /
• 2018

The purpose of this work is to solve the problem of representing an entire video using Convolutional Neural Network (CNN) features for human action recognition. Recently, due to insufficient GPU memory, it has been difficult to take the whole video as the input of the CNN for end-to-end learning. A typical method is to use sampled video frames as inputs and corresponding labels as supervision. One major issue of this popular approach is that the local samples may not contain the information indicated by the global labels and sufficient motion information. To address this issue, we propose a binary hashing method to enhance the local feature extractors. First, we extract the local features and aggregate them into global features using maximum/minimum pooling. Second, we use the binary hashing method to capture the motion features. Finally, we concatenate the hashing features with global features using different normalization methods to train the classifier. Experimental results on the JHMDB and MPII-Cooking datasets show that, for these new local features, binary hashing mapping on the sparsely sampled features led to significant performance improvements.

• Journal of Ship and Ocean Technology
• /
• v.8 no.3
• /
• pp.26-39
• /
• 2004

A vessel/mooring/riser coupled dynamic analysis program in time domain is developed for the global motion simulation of a turret-moored, tanker based FPSO designed for 6000-ft water depth. The vessel global motions and mooring tension are simulated for the non-parallel wind-wave-current 100-year hurricane condition in the Gulf of Mexico. The wind and current forces and moments are estimated from the OCIMF empirical data base for the given loading condition. The numerical results are compared with the OTRC(Offshore Technology Research Center: Model Basin for Offshore Platforms in Texas A&M University) 1:60 model-testing results with truncated mooring system. The system's stiffness and line tension as well as natural periods and damping obtained from the OTRC measurement are checked through numerically simulated static-offset and free-decay tests. The global vessel motion simulations in the hurricane condition were conducted by varying lateral and longitudinal hull drag coefficients, different mooring and riser set up, and wind-exposed areas to better understand the sensitivity of the FPSO responses against empirical parameters. It is particularly stressed that the dynamic mooring tension can be greatly underestimated when truncated mooring system is used.

• International Journal of Advanced Culture Technology
• /
• v.11 no.1
• /
• pp.296-305
• /
• 2023

The purpose of this study is to measure and compare the balance of motion between the left and right using a wearable sensor during upper limb exercise using an exercise equipment. Eight participants were asked to perform upper limb exercise using exercise equipment, and exercise data were measured through IMU sensors attached to both wrists. As a result of the PCA test, Euler Yaw(Left: 0.65, Right: 0.75), Roll(Left: 0.72, Right: 0.58), and Gyro X(Left: 0.64, Right: 0.63) were identified as the main components in the Butterfly exercise, and Euler Pitch(Left: 0.70, Right 0.70) and Gyro Z(Left: 0.70, Right: 0.71) were identified as the main components in the Lat pull down exercise. As a result of the Paired-T test of the Euler value, Yaw's Peak to Peak at Butterfly exercise and Roll's Mean, Yaw's Mean and Period at Lat pull down exercise were smaller than the significance level of 0.05, proving meaningful difference was found. In the Symmetry Index and Symmetry Ratio analysis, 89% of the subjects showed a tendency of dominant limb maintaining relatively higher angular movement performance then non-dominant limb as the Butterfly exercise proceeds. 62.5% of the subjects showed the same tendency during the Lat pull down exercise. These experimental results indicate that meaningful difference at balance of motion was found according to an increase in number of exercise trials.