Earticle

In this paper, we design and implement a distributed, moving objects management system for processing locations and sensor data from smart black boxes. The proposed system is designed and implemented based on Apache Kafka, Apache Spark & Spark Streaming, Hbase, HDFS. Apache Kafka is used to collect the data from smart black boxes and queries from users. Received location data from smart black boxes and queries from users becomes input of Apache Spark Streaming. Apache Spark Streaming preprocesses the input data for indexing. Recent location data and indexes are stored in-memory managed by Apache Spark. Old data and indexes are flushed into HBase later. We perform experiments to show the throughput of the index manager. Finally, we describe the implementation detail in Scala function level.

Generally, existing parallel main-memory spatial index structures to avoid the trade-off between query freshness and CPU cost uses light-weight locking techniques. However, still, the lock based methods have some limits such as thrashing which is a well-known problem in lock based methods. In this paper, we propose a distributed index structure for moving objects exploiting the parallelism in multiple machines. The proposed index is a lock free multi-version concurrency technique based on the D-Stream model of Spark Stream. The proposed method exploits the multiversion nature of D-Stream of Spark Streaming.

Comprehension of physical events in terms of cause and effect is fundamental for making sense of and dealing successfully with changes in the dynamic physical world. Previous research has demonstrated that the causal structure of the world can, in some cases, be directly perceived: When two billiard balls collide, observers perceive that the action of one ball caused the other's motion, merging two motion events into a unitary percept. The current study explored whether such casual interpretations can contribute to resolving low-level ambiguities in motion perception. We used a bistable apparent motion display, a motion quartet, which can lead to the perception of either horizontal or vertical motion, and tested the effects of “context objects” which moved in such a way that motion targets appeared to collide with them in either horizontal or vertical dimension. Our results show that contextual motion implying a Michotte-style launch can strongly bias observed motion correspondence, consistent with physical regularities of mechanical causality in a postdictive way. It suggests that the perception of causality is an earlier and more pervasive phenomenon than previously understood, and in fact can influence the perception of motion itself.

본 논문에서는 미지의 환경에서 알려지지 않는 이동 물체를 추적하는 기법을 제안하였다. 이 기법은 기존의 JPDA 방법에 비해서 컴퓨터 계산시간이 매우 단축되는 장점이 있다. 제안된 기법은 컴퓨터 시뮬레이션을 통하여 추적의 특성을 확인하였다. 본 기법은 파라메터의 선정이 매우 중요한데 본 논문에서는 파라메터 선정을 위한 relaxation 기법을 함께 제안하여 실제 응용에 문제가 없는 것을 확인하였다.

In this paper, I have presented a new approach which can track moving objects in unknown environments. This scheme is important in providing a computationally feasible alternative to complete enumeration of JPDA which is intractable. I have proved that given an artificial measurement and track's configuration, proposed scheme converges to a proper plot in a finite number of iterations. In this light, even if the performance is enhanced by using the relaxation, I also note that the difficulty in tuning the parameters of the relaxation scheme is critical aspect of this suggestion.

In this study, a novel method for dimension measurement of large-scale moving objects using stereo camera with 2-degree of freedom (2-DOF) mechanism is presented. The proposed method utilizes both the advantages of stereo vision technique and the enlarged visibility range of camera due to 2-DOF rotary mechanism in measuring large-scale moving objects. The measurement system employs a stereo camera combined with a 2-DOF rotary mechanism that allows capturing separate corners of the measured object. The measuring algorithm consists of two main stages. First, three-dimensional (3-D) positions of the corners of the measured object are determined based on stereo vision algorithms. Then, using the rotary angles of the 2-DOF mechanism the dimensions of the measured object are calculated via coordinate transformation. The proposed system can measure the dimensions of moving objects with relatively slow and steady speed. We showed that the proposed system guarantees high measuring accuracy with some experiments.

This study presents and algorithm and related techniques which could satisfy the important properties of check weighers and conveyor scales. The algorithm of Recursive Least Squares Regression is applied for the weighing system simulated as a dynamic model of the second order. Using the model and the algorithm, model parameters and then the mass being weighed can be determined from the step input. The performance of the algorithm was tested on a check weigher. Discussions were extended to the development of noise reduction techniques and to the lagged introduction of objects on the moving plate. It turns out that the algorithm shows several desirable features suitable for real-time signal processing with a microcomputer, which are high precision and stability in noisy environment.

.

본 논문은 Snake 모델을 이용하여 동영상에서 주위 환경 변화에 적응 가능한 다중 이동 객체 추적 시스템을 제안하였다. Snake 모델은 배경이 복잡한 영상에 대해선 객체의 윤곽선을 정확히 표현하지 못하므로 영상분할 시 초기 위치에 따라 민감하게 영향을 받는다. 제안된 시스템은 프레임간의 차(difference)영상을 이용하여 배경영상을 획득하고, 픽셀의 인접성을 조사하여 객체를 분할하고 위치 특징 값을 구하며, 분할된 특징 값들을 Snake모델의 초기 위치 값으로 부여함으로써 초기 위치 값에 민감한 Snake 모델을 개선하였다 또한 본 시스템은 복잡한 배경 영상을 단순화하고, Snake를 이루는 각 정점들을 객체의 위치로 놓이게 함으로써 탐색 공간을 줄였다. 30fps로 저장된 AVI파일을 적용함으로써 다중 이동차량 추적 시스템으로의 응용 가능함을 보였다.

This paper proposes a multiple moving objects tracking system which is adaptable itself to circumstances. Snake model is sensitive to the start position value because it does not accurately express contours of objects in complex image. It can be improved as the proposed system gets background images by using difference images, segments objects using neighborhood pixels and assesses the position feature values acquired on the start position value to deformable Snake model. And also the system can simplify complex background images and reduce search regions by the constituent points of a Snake laid in Positions of object. It is showed that the proposed system can be appBied to multiple moving vehicle racking systems by the experimental results of 30fps AVI file.

본 논문에서는 실내 조명 변화 상황에 강인한 이동 물체 검출 알고리즘을 제안한다. 기존 알고리즘이 단일 픽셀 혹은 인접 픽셀들의 강도에 기반을 두기 때문에 빛 변화 상황에서 오검출률이 급격하게 증가하는 단점이 있다. 이러한 문제를 해결하기 위하여 local binary pattern 알고리즘을 기반으로 대칭되는 위상에 존재하는 인접 픽셀 간의 그래디언트 벡터를 통해 영상을 변환하기 때문에 광량과 조명의 변화에 강인한 그래디언트 벡터 기반 영상 변환(i.e. gradient-vector based image transform) 알고리즘과 유사성 기반 배경 모델링 알고리즘인 multi-channel kernel fuzzy correlogram(MKFC)을 결합하여 빛 또는 조명 변화 변화에 강인한 새로운 이동 물체 검출 알고리즘을 제안한다. 끝으로, 시뮬레이션을 통하여 급격한 조명 변화 상황에서도 제안된 알고리즘이 기존 알고리즘보다 낮은 오검출률을 보임을 검증한다.

In this paper, we propose a robust moving object detection algorithm against indoor light illumination conditions. Since the conventional algorithm is based on the intensity of pixels, its false alarm rate may increase in such indoor light conditions. On the other hand, the proposed algorithm uses information of gradient vector of pixel intensity variance based on local binary patterns (i.e., gradient-vector based image transform), resulting in insensitivity to light change situations. Finally, the performance of the proposed algorithm is verified in terms of false alarm rates through simulation results.

기존 증강현실 콘텐츠 경우 고정된 마커를 빠르게 이동시키면 가시화에 끊김 현상이 발생하게 된다. 따라서 동적 마커를 사용할 경우에도 콘텐츠가 끊어짐 없이 사용자에게 제공되는 방법이 요구된다. 카메라에 입력된 한 장 의 이미지 내에 두 개 이상의 마커가 존재할 경우 기존의 이미지 기반 마커와 SLAM(Simultaneous Localization & Mapping) 방식을 통해서는 각각의 마커를 동시에 추적할 수 없다. 본 논문은 각 마커 위에 정합된 객체들 간의 상호작용은 불가능하다는 점을 극복함은 물론, 빠르게 움직이는 마커를 실시간으로 추적하여 그 위에 원하는 객체를 정확하게 증강 가시화하는 방법론을 제안한다. 이를 위해, 주행형 로봇 과 시범 콘텐츠의 가상로봇을 동기화하여, 주행형 로봇 상에 콘텐츠가 가시화되도록 하였다. 그리고 카메라 한 대로 다중 동적객체를 추적하여 서로 상호작용하는 기술을 제안하였다. 결과적으로 가상로봇과 실제로봇을 연동하여 상호 작용하도록 함으로써 실시간 동적객체 추적 및 가시화 기술의 유용성을 검증하였다.

AR contents, if markers to be tracked move fast, show flickering while updating images captured from cameras. Conventional methods employing image based markers and SLAM algorithms for tracking objects have the problem that they do not allow more than 2 objects to be tracked simultaneously and interacted with each other in the same camera scene. In this paper, an improved SLAM type algorithm for tracking dynamic objects is proposed and investigated to solve the problem described above. To this end, method using 2 virtual cameras for one physical camera is adopted, which makes the tracked 2 objects interacted with each other. This becomes possible because 2 objects are perceived separately by single physical camera. Mobile robots used as dynamic objects are synchronized with virtual robots in the well-designed contents, proving usefulness of applying the result of individual tracking for multiple moving objects to augmented visualization of objects.

위치기반 서비스란 휴대폰과 같은 모바일 단말기 속에 위성 항법장치(GPS)와 연결되는 칩을 부착하여 위치추적 서비스, 공공안전 서비스, 위치기반 정보 서비스 등 위치와 관련된 각종 정보를 제공하는 서비스를 일컫는다. 그리고 클라우드 컴퓨팅이란 개인용 컴퓨터 또는 기업의 서버에 개별적으로 저장해 두었던 자료와 소프트웨어들을 클라우드 클러스터로 구축하여 필요할 때 PC나 휴대폰 같은 각종 단말기를 이용하여 원격 작업을 수행할 수 있는 환경을 의미한다. 본 논문에서는 구글의 연구로부터 시작된 클라우드 컴퓨팅 기술에 주목하고 클라우드 컴퓨팅 환경을 위한 대용량 이동 객체의 정보를 저장하고 질의하는 시스템에 대하여 연구하였다. 그리고 위치 기반 서비스를 위한 대용량 이동객체에 대하여 효율적인 질의를 수행할 수 있는 저장 구조를 제안하였다. 본 논문에서 제안한 방법을 적용한 시스템의 효율성을 증명하기 위해 저장 및 질의 성능을 다양한 측면에서 실험하고, 실험 결과를 다른 분산처리 시스템과의 비교하여 성능을 입증하였다.

The location-based service is referred to as the service for providing various types of information regarding such as the location-tracing service, the public-safety service and the location-based information service by attaching chips connected with the GPS to the mobile phone. In addition, cloud computing means the environment which makes it possible to carry out remote operations by establishing cloud clusters with the data and software which have been individually stored in personal computer or corporate servers and using devices such as personal computers and cellular phones whenever it is necessary. Throughout this paper, the cloud computing technology which originates from the research by Google the system has been focused on and the system for storing a large volume of information for moving object and making query for cloud computing environment has been studied. Besides, the storing structure for executing efficient query regarding the high-capacity mobile devices for the location-based service has been suggested. In order to prove the efficiency of the system with the application of the method suggested in this paper, the storage and qualitative functions have been experimented in various aspects and compared with other distributed processing system's performance.

최근 도로상의 비디오 영상에서 다양한 저차원 정보를 바탕으로 자동 차량 흐름 파악과 사고 탐지에 관해 관심이 높아지고 있다. 본 논문에서는 CCTV 동영상 교통관리시스템을 이용하여 자동으로 차량의 흐름을 파악하고 이를 이용하여 교통체증과 더 나아가 사고탐지를 위한 알고리즘과 그 응용에 대해 연구하였다. 이를 위하여 우선 움직임 객체를 시공간 관계 모델링을 통해 차량의 실제 궤적과 매핑하고 이를 통해 차량의 흐름 파악에 사용하였다. 또한 시공간 관계 모델링과 차량의 실제 궤적간 매핑을 위해 TSR (Tangent Space Representation) 알고리즘을 사용하였다. 또한, 객체의 움직임 추출을 위해 먼저 차(Differece)영상을 이용하여 움직임 객체 추출을 하였고, 이를 통하여 객체의 움직임 트래킹(Tracking)을 하고 각 객체에 번호를 부여하여 동시에 여러 객체를 인식시키고 이를 저장하였다. 이를 통하여 의미적 움직임 객체를 인식하고, 차량 흐름을 파악하는 어플리케이션을 구현하였다. 본 연구를 통해 기본적인 움직임 객체에 대한 의미적인 결과를 얻을 수 있었고, 나아가 CCTV를 이용한 자동 사고 탐지에 관한 연구로 확장시킬 수 있을 것으로 기대된다.

Recently, there are interested in the automatic traffic flowing and accident detection using various low level information from video in the road. In this paper, the automatic traffic flowing and algorithm, and application of traffic accident detection using traffic management systems are studied. To achieve these purposes, the spatio-temporal relation models using topological and directional relations have been made, then a matching of the proposed models with the directional motion verbs proposed by Levin’s verbs of inherently directed motion is applied. Finally, the synonym and antonym are inserted by using WordNet. For the similarity measuring between proposed modeling and trajectory of moving object in the video, the objects are extracted, and then compared with the trajectories of moving objects by the proposed modeling. Because of the different features with each proposed modeling, the rules that have been generated will be applied to the similarity measurement by TSR (Tangent Space Representation). Through this research, we can extend our results to the automatic accident detection of vehicle using CCTV.

Researches for NN(nearest neighbor) query which is often used in LBS system, have been worked. However. Conventional NN query processing techniques are usually meaningless in moving object management system for LBS since their results may be invalidated as soon as the query and data objects move. To solve these problems, in this paper we propose a new nearest neighbor query processing technique, called CTNN, which is possible to meet continuous trajectory nearest neighbor query processing. The proposed technique consists of Approximate CTNN technique which has quick response time, and Exact CTNN technique which makes it possible to search accurately nearest neighbor objects. Experimental results using GSTD datasets shows that the Exact CTNN technique has high accuracy, but has a little low performance for response time. They also shows that the Approximate CTNN technique has low accuracy comparing with the Exact CTNN, but has high response time.

영상보안은 CCTV 카메라, 디지털 저장장치, 영상분석 및 인식기술 등의 발전과 시장 확대로 인하여 실시간 감시 시스템이나 지능형 영상보안 시스템 등의 보다 진화된 기술의 개발을 요구하고 있다. CCTV 카메라 시스템에서 회전이 가능한 PTZ 카메라는 줌인 기능을 가지고 있으므로 물체에 대한 정밀한 영상을 획득할 수 있으나 사각 지대가 발생할 수 있으며 동시에 두 개 이상의 이동하는 물체에 대한 감시가 불가능한 단점 을 가지고 있었다. 본 연구는 지능형 동체추적 CCTV 시스템 및 영상 감시 방법에 관한 것으로 전체 영역을 감시하는 하나의 고정 카 메라와 두 개 이상의 PTZ 카메라를 사용하여 광범위 영역에서 동시에 이동하는 다수의 물체를 정밀하게 촬영할 수 있는 지능형 영상 보안 시스템을 제안한다.

Due to the development and market expansion of image analysis and recognition technology, video security such as CCTV cameras and digital storage devices, are required for real-time monitoring systems and intelligent video security systems. This includes the development of more advanced technologies. A rotatable PTZ camera, in a CCTV camera system, has a zoom function so you can acquire a precise picture. However it can cause blind spots, and can not monitor two or more moving objects at the same time. This study concerns, the intelligent tracking of multiple moving objects, CCTV systems, and methods of video surveillance. An intelligent video surveillance system is proposed. It can accurately shoot broad areas and track multiple objects at the same time, much more effectively than using one fixed camera for an entire area or two or more PTZ cameras.

영상취득 장치를 이용한 지능화된 감시 장치의 개발 기술 또한 발전하고 있다. 비교적 고속으로 움직이는 객체를 탐지해야 하는 분야에서 무엇보다 중요한 것은 배경영상 갱신에 대한 부하를 효과적으로 줄여서 실시간적으로 갱신할 수 있어야 하는데 현재 범용 컴퓨터 능력으로는 질감 등을 특징으로 추출하는 방법 등은 대부분 연산처리의 부하 때문에 적용상의 한계가 있다. 본 논문에서는 적어도 초당 30프레임의 카메라 영상에서 주행 중인 자동차와 같이 고속으로 움직이는 객체를 탐지하는 응용영역에서 실시간으로 배경 영상을 갱신하는 알고리즘을 제시하고, 실제 입력영상에서 객체 영역을 추출하는 시험을 통해 성능을 분석하였다.

A developed skill of an intelligent CCTV is also advancing by using its Image Acquisition Device. The most important part in the field of detecting comparatively fast moving objects is to effectively reduce the loads on updating the background image in order to achieve real－time update. However, the ability of the current general－purpose computer extracting the texture as characteristics has limits in application mostly due to the loads on processes. In this thesis, an algorithm for real－time updating the background image in an applied area such as detecting the fast moving objects like a driving car in a video of at least 30 frames per second is suggested and the performance is analyzed by a test of extracting object region from real input image.

Many indexing methods were proposed so that process moving object efficiently. Among them, indexing methods like the 3D R-tree treat temporal and spatial domain as the same. Actually, however, both domain had better process separately because of difference in character and unit. Especially in this paper we deal with limited region such as indoor environment since spatial domain is limited but temporal domain is grown. In this paper we present a novel indexing structure, namely STS-tree(Separation of Time and Space tree), based on limited region. STS-tree is a hybrid tree structure which consists of R-tree and one-dimensional TB-tree. The R-tree component indexes static object and spatial information such as topography of the space. The TB-tree component indexes moving object and temporal information.

일반적으로 게임 디자인 시에 웹과 같은 매체를 통해 구하거나 디자이너들이 디자인 툴로 제작한 3차원 기하 모델은 polygon soup 형태가 대부분이다. 따라서 이러한 polygon soup 모델은 일반적으로 완전한 매시로 가정해서 적용하는 여러 기법들을 적용할 수 없다. 이러한 문제를 해결하고 보다 매끈한 곡면을 얻기 위해서 MLS(Moving Least Squares) 방법을 점 단위가 아니라 삼각형 면 단위로 확장한 적분형 MLS 기법이 제안되었다. 그러나 이 기법은 본질적으로 전역(global) 계산의 한계로 인한 계산 속도의 한계가 불가피한 특징이 있었다. 본 논문에서는 분석적 해를 바탕으로 한 전역 계산 속도를 보다 가속화하기 위한 GPU 기반 병렬 기법을 제안하고 또한 기존 논문에서 충분히 논의하지 않았던 적분형 MLS 기법과 일반 MLS 기법의 차이점을 설명한다. 특히 GPU를 통한 가속 결과, 연산 정밀도의 감소없이 CPU 코어 1개로 계산하는 경우보다 평균 250배의 속도 향상을 얻을 수 있었다.

Most 3D models which are easily available on the Internet or designed by designers are polygon soups - they do not have topology information; some triangles are missing; or there exist some T-junctions. Due to this issue, it is often impossible to apply advanced geometry techniques to those polygon soups because most advanced methods assume that the 3D models are perfect manifold meshes. To address this problem, an integral MLS method which calculates integration, not summation over sample points to generate smooth implicit surfaces. However, those methods basically require global calculations for all the triangles so that it is very hard to accelerate the operations only with one CPU core. In this paper, we present a GPU-based parallelization technique for the analytic approach of the integral MLS method. Also, we discuss the difference between integral MLS and conventional one which was not fully explained in other papers. In our experiment, using GPU, we achieve that 250-fold speedup over the single CPU core.

With the development of the Internet of Things (IoT) and indoor positioning technologies, how to manage the rapidly increasing indoor moving object data is a research topic. The fundamental issue is to consider the representation of complicated indoor space as well as the moving object. In this paper, we aim to propose a complete indoor space model that can sup-port various indoor applications. The proposed model divides each indoor element into multi-granularity grids and introduces grids as indoor primitive geometry for the representation of connectivity between indoor elements and location of moving objects. The advantage of this approach is that it can simultaneously represent the connective relationship and geometric information of indoor elements, and give more accurate description of indoor distance and direction information. Several queries of indoor applications under the shopping mall scenar-io are employed to illustrate the completeness and robustness of our model.