AMR를 이용한 가변적인 관심 대상 객체 추적을 위한 딥러닝 기반 프레임워크

곽정훈,양견모,구재완,서갑호 제어·로봇·시스템학회 2022 제어·로봇·시스템학회 논문지 Vol.28 No.12

AMR(Autonomous Mobile Robot) is being used to improve working environment through collaboration such as transporting goods between workers. For collaboration such as transporting goods, AMR tracks the workers and carries out goods transport. Object tracking is possible based on a deep learning model trained using big data, built as an object to be tracked. When the worker changes frequently, such as in a work environment, there is a problem in that big data construction and deep learning model learning are required whenever an object to be tracked is changed. There is a need for a method for tracking objects that change frequently while providing small amounts of data. This paper proposes a deep learning-based framework for tracking changeable object. An object to be tracked, such as a worker, is defined as a ToI (Target-of-Interest) object. The proposed framework utilizes a two-stage deep learning model to track a changeable ToI object. In the deep learning model of the first stage, an object of the same type as the ToI object is tracked. In the deep learning model of the second stage, the ToI object is found among the objects being tracked. The position of the ToI object is transformed into the coordinate system of the AMR so that the AMR can track the ToI object. In the experiment, the results of tracking the ToI object by using the proposed method were verified. When tracking ToI objects with a single-stage deep learning model with a small amount of data, the accuracy of tracking the ToI objects decreased according to the amount of data. In the case of the proposed method, the tracking of the ToI object was not affected by the amount of data. .

[컴퓨터지능 및 지능시스템] DTW 알고리즘을 이용한 다중 이동 객체 추적 방법

왕교선(Kyo Sun Wang),주영훈(Yung Hoon Joo) 대한전기학회 2019 전기학회논문지 Vol.68 No.5

In this paper, we propose a method to track multiple moving objects using DTW algorithm for images received from CCTV Cameras. The proposed method first extracts a moving object using GMM background modeling to extract a moving object from the input image, and then removes and extends the shadow and noise of the extracted moving object using morphology. The moving object thus obtained is recognized by using the labeling method, and the labeling is merged using the morphology to identify the problem that the recognized moving object is divided into several labels. Then, when the recognized moving object is detected in the designated ROI, the HSV and RGB values of the moving object are extracted and the extracted data is sequentially stored in the DTW data base. Next, we propose a method of classifying the moving object as similar to the DTW by judging the similarity between the stored DTW DB and the moving object. The moving objects classified by the proposed method are continuously compared and tracked. If it is judged that the moving objects to be tracked are overlapped, it is possible to compare the remaining objects. It also suggests how tracking can be done if the moving object is unwrapped. Finally, the applicability of the proposed method is verified through various experiments.

Detection of Mobile Object in Workspace Area

Shah, H.N.M,Rashid, M.Z.A,Abdollah,M.F,Kamarudin, M.N,Kamis, Z,Khamis, A 보안공학연구지원센터 2016 International Journal of Signal Processing, Image Vol.9 No.4

This paper introduces the detection of mobile object in intelligent space robot application. There are three major algorithms, namely object detection, object classification and object tracking. The core of the detection of mobile object comprise of two processes: offline and online. An offline process consists of the training of the model using deference input sources that depend on the application. An online process consists of the matching process and the result of the object poses. The main idea of object classification is to classify into two categories depending on the dimension of object, mobile object and non-mobile object. By using an offline and an online process the whole process becomes faster because there only have object classification and object tracking involved in real time. The positions of the mobile object are represented by symbol X with difference colors for easy comparison with non-mobile object. One of the unique advantages mentioned in this paper, the detection of mobile object only uses image processing that are generated by the algorithms without additional sensor like sonar or IR sensor.

춤에서 오브제(Object)의 다면적 활용과 역할 연구: 사샤 발츠(Sasha Waltz)와 윌리엄 포사이드(William Forsythe)의 설치 작품을 중심으로

박서영 대한무용학회 2020 대한무용학회논문집 Vol.78 No.5

This study aims to research major roles of an object in dance by presenting specific examples of strategies showing how an object is utilized in contemporary dance. Nowadays pieces of contemporary dance often show a variety of objects which entered the space of art. In pieces of art, objects are playing so many different roles that they cannot even be described collectively. The concept of an 'object' which means an object to be recognized, by definition, as opposed to a subject is constantly redefined by changes in the structure and meaning of things in different times. In particular, in the exhibition of dance where even a human body, a subject that seemed impossible to be objectified, is naturally experienced as an object, audience can recognize the broad spectrum of an object from 'a subject to an object'. Such wide spectrum makes an object work as a mediator among visual art, play and dance. In the field of dance, an object is presenting various strategies reformulating the concept of a 'body', 'space' and 'time' and broadening the scope of dance. .

중학교 한문과 “한문지식영역”에서 한문문법의 문제 -빈어와 보어의 설정문제-

정순영 한국한문교육학회 2008 한문교육논집 Vol.31 No.-

This paper studied Object and definition of Complement concepts of Chinese grammar as shown the 7th middle school curriculum of Korean Classical Chinese Education as amended in 2007. The curriculum of Chinese Education was represented validity of amend which is fitting in general system in the grammar as grammar parts, contents, naming and concepts of terminologies, also naming of terminologies was represented well. But rule of concepts of Object and Complement as shown the middle school curriculum have been ambiguous, superficial. Also, example sentence which can be chaos in the education on site to novice. A chaos of Object and Complement of the curriculum of Korean Classical Chinese Education as amended came from ambiguity of statements. So, this paper proposed to correct rule of concepts as follows ; -Object is 'Predicate represent behavior or existence and Object became the object' in the past but Object is 'Predicate represent recognition, comparison, designation, thinking and Object became the object in the as amended. -Complement is 'Predicate represent behavior or condition and Comple -ment support meaning fully in the past but Complement is 'Predicate represent condition or equal connection and Complement support meaning of condition, equal connection in the as amended. As above, this paper proposed to specify rule of Predicate of Object and added to recognition, comparison and designation in the action items. Also, omit 'behavior' item in the statement of Predicate of Object and rule of concepts proposed to move definition items 'existence or possess -ion' in the rule of Object and proposed to add to describe be verb which represent 'equal connection'. Nowadays, Chinese grammar is not Object for using in language's life, it is tool for reading comprehension which is describing systematic pheno -mena school as recorded. Therefore School grammar must be described the basis on Scholarship grammar. Above all, considering to school on site, it need to develop for learning which could be explained and under -stood. 본 논문은 개정 7차 한문과 중학교 교육과정해설서에 나타난 “한문지식영역”에서 한문문법의 빈어와 보어개념설정문제에 관한 연구다. 개정교육과정은 문법범위나 내용, 용어의 명칭과 개념 등 문법방면에 있어서 타 교과와 보편적 체계에 맞추어 개정의 타당성을 잘 반영하고 있으며, 용어의 명칭도 한문의 어법적 특징을 잘 반영하고 있다. 그러나 개정중학교교육과정에 보이는 빈어와 보어의 영역구분 그리고 개념규정이 불분명하고, 피상적으로 되어있으며, 예시문도 현장교육에서 초학습자에게 혼란을 가져올 수 있는 예문을 제시하고 있다. 개정교육과정의 빈어와 보어의 혼란은 개념규정을 모호하게 진술한데서 비롯되었으므로 필자는 빈어와 보어의 개념규정을 아래와 같이 수정할 것을 제언 한다. 빈어-‘서술어는 동작이나 행위 또는 존재나 소유를 나타내고, 빈어는 그 대상이 된다.’에서 빈어-‘서술어는 동작이나 행위(인식, 비교, 칭위, 사유)를 나타내고, 빈어는 그 대상이 된다.’ 보어-‘서술어는 동작, 행위, 상태 등을 나타내고, 보어는 서술어를 보충하여 부족한 뜻을 완전하게 해준다.’에서 보어-‘서술어는 상태나 동일관계를 나타내고, 보어는 주어의 상태 또는 존재, 소유나 동일관계의 의미를 보충하여 준다. 위에서 볼 수 있는 바와 같이, 빈어의 서술어 규정을 더욱 구체화하여 행위 항목에 ‘인식, 비교, 칭위, 사유’ 등의 항목을 추가할 것을 제언한다. 또한 보어의 서술어 진술에서 ‘동작 행위’ 항목을 생략하고, 빈어의 규정에서 ‘존재나 소유를 나타내고’ 의 항목을 보어의 개념규정 항목으로 이동하여 진술할 것을 제언한다. 그리고 ‘동일관계’를 나타내는 계사동사를 추가설정 할 것을 제언한다. 서술어의 지배대상이 되지 않는 비 동작성의 상태, 존재, 소유, 연계를 나타내는 서술어 다음에 오는 성분은 빈어의 위치에 있지만, 지배관계를 갖지 않고 주어의 상태 또는 동일관계를 나타내므로 보어로 규정하는 것이 타당하다. 학교 한문문법은 지금 언어생활에서 사용하기 위한 대상은 아니다. 한문문법은 기록되어 있는 문언문에 대한 체계적 현상을 기술하는 것으로 독해를 위한 수단이다. 그러므로 학교문법은 학문문법 체계를 근거로 기술되어야 하지만, 무엇보다도 학교 현장을 고려하여 학습자에게 쉽게 설명되고 이해될 수 있는 학습 방법의 개발이 절실하다.

Object Motion Tracking using a Moving Direction Estimate and Color Updates

Samuel Henry Chang,심덕선,김희영,최광남 제어·로봇·시스템학회 2012 International Journal of Control, Automation, and Vol.10 No.1

This paper presents a direction detection and tracking object color update algorithm used to track moving objects that change colors. Different from traditional color-based tracking methods, which use an initial color distribution in order to track objects as long as the object carries the full or partial initial color, this method introduces a color update method used to quickly find the new object color in a new location if the object changes its color partially or completely; the updated color is then used to locate the object. In our algorithm, an initial color pattern is used to track an object using the color. During the tracking, an object’s new location is at first estimated and then used to detect any color change. If the color has changed, a new color pattern is updated based on the changes in the previous color distribution, and then the new color pattern is used to calculate the current location of the object. This algorithm utilizes the property that the movement of an object can be estimated either by using the object’s shadow or by background subtraction. The implementation of our algorithm results in an effective real-time object tracking. The validity of the approach is illustrated by the presentation of experiment results obtained using the methods described in this paper.

정확히 재가중되는 온라인 전체 에러율 최소화 기반의 객체 추적

장세인,박충식 한국지능정보시스템학회 2019 지능정보연구 Vol.25 No.4

Object tracking is one of important steps to achieve video-based surveillance systems. Object tracking is considered as an essential task similar to object detection and recognition. In order to perform object tracking, various machine learning methods (e.g., least-squares, perceptron and support vector machine) can be applied for different designs of tracking systems. In general, generative methods (e.g., principal component analysis) were utilized due to its simplicity and effectiveness. However, the generative methods were only focused on modeling the target object. Due to this limitation, discriminative methods (e.g., binary classification) were adopted to distinguish the target object and the background. Among the machine learning methods for binary classification, total error rate minimization can be used as one of successful machine learning methods for binary classification. The total error rate minimization can achieve a global minimum due to a quadratic approximation to a step function while other methods (e.g., support vector machine) seek local minima using nonlinear functions (e.g., hinge loss function). Due to this quadratic approximation, the total error rate minimization could obtain appropriate properties in solving optimization problems for binary classification. However, this total error rate minimization was based on a batch mode setting. The batch mode setting can be limited to several applications under offline learning. Due to limited computing resources, offline learning could not handle large scale data sets. Compared to offline learning, online learning can update its solution without storing all training samples in learning process. Due to increment of large scale data sets, online learning becomes one of essential properties for various applications. Since object tracking needs to handle data samples in real time, online learning based total error rate minimization methods are necessary to efficiently address object tracking problems. Due to the need of the online learning, an online learning based total error rate minimization method was developed. However, an approximately reweighted technique was developed. Although the approximation technique is utilized, this online version of the total error rate minimization could achieve good performances in biometric applications. However, this method is assumed that the total error rate minimization can be asymptotically achieved when only the number of training samples is infinite. Although there is the assumption to achieve the total error rate minimization, the approximation issue can continuously accumulate learning errors according to increment of training samples. Due to this reason, the approximated online learning solution can then lead a wrong solution. The wrong solution can make significant errors when it is applied to surveillance systems. In this paper, we propose an exactly reweighted technique to recursively update the solution of the total error rate minimization in online learning manner. Compared to the approximately reweighted online total error rate minimization, an exactly reweighted online total error rate minimization is achieved. The proposed exact online learning method based on the total error rate minimization is then applied to object tracking problems. In our object tracking system, particle filtering is adopted. In particle filtering, our observation model is consisted of both generative and discriminative methods to leverage the advantages between generative and discriminative properties. In our experiments, our proposed object tracking system achieves promising performances on 8 public video sequences over competing object tracking systems. The paired t-test is also reported to evaluate its quality of the results. Our proposed online learning method can be extended under the deep learning architecture which can cover the shallow and deep networks. Moreover, online learning methods, that need the exact reweighting pro... 영상 기반의 보안 시스템의 증가함에 따라 각 용도마다 다른 다양한 객체들에 대한 처리들이 중요해지고 있다. 객체 추적은 객체 인식, 검출과 같은 작업들과 함께 필수적인 작업으로 다뤄진다. 이 객체 추적을 달성하기위해서 다양한 머신러닝이 적용될 수 있다. 성공적인 분류기로써 전체 에러율 최소화(total-error-rate minimization) 기반의 방법론이 사용될 수 있다. 이 전체 에러율 최소화 기반의 방법론은 오프라인 학습을 기반으로 하고 있다. 객체 추적은 실시간으로 처리하며 갱신해야하는 것이 필수적이므로 온라인 학습(online learning)을 기반으로 하는 것이 적합하다. 온라인 전체 에러율 최소화 방법론이 개발되었지만 점근적으로 재가중되는(approximately reweighted) 작업이 포함되어 에러를 누적시킬 수 있다는 단점이 있다. 본 논문에서는 정확하게 재가중되는(exactly reweighted) 방법론을 제안하면서 온라인 전체 에러율 최소화가 달성되었다. 이 제안된온라인 학습 방법론을 객체 추적에 적용하여 총 8개의 데이터베이스에서 다른 추적 방법론들 보다 좋은 성능이달성되었다.

“A+着”的賓語

이아형 ( Lee Ahyoung ) 한국중국언어학회 2016 중국언어연구 Vol.0 No.66

1441個形容詞中能帶“着”的爲941個義項,其中將近半數的形容詞能帶賓語。但實際情況看,其賓語之間的特征大有不同。所以我們將“A+着”的賓語分爲三種:特殊賓語、使動賓語、抽象賓語。特殊賓語再可以分爲三種:“整體-部分”類、介賓類以及存在類。使動賓語可以分爲兩種:“整體-部分”類以及“原因-結果”類。抽象賓語基本上是施事和受事的關系。從與時間詞共現的角度看,特殊賓語只與恒量時間詞一起共現,使動賓語可以與恒量時間詞以及變量時間詞一起共現,與變量時間詞一起共現的比例更高,抽象賓語與時間詞的共現?自由。從此可見,雖然形容詞與“着”相結合后比形容詞更容易帶賓語,但是從賓語的內部情況看,“A+着”的賓語與動詞的賓語的屬性截然不同. 941 meaning of the 1441 meanings of the adjectives can bring ‘zhe`, nearly half of these adjectives can take an object. But the actual situation, the feature between the object are different. So we divide the object of “Adj+zhe” into three kinds, such as special object, causative object, abstract object. The special object can be divided into three kinds, for example, the class of ‘the whole-Part`, the class of Preposition object, the class of existential sentence. The causative object can be divided into two types, for example, the class of ‘the whole - Part`, the class of ‘Cause - result`. The abstract object basically is the relationship between the agent and the patient. From the perspective of the time word co-occurrence, the special object is only with constant time words together. The causative object is with constant time words and variable time words together, and the proportion of time words with variable time are higher than constant time words. Co-occurrence of The abstract object and the time words are free. From now on, even though after the adjective combines with ‘zhe`, can be easy to take object, but from the inside of the object,the object of ‘Adj+zhe` is different from the object of the verb.

An Object-Oriented Design Methodology Based on Object Service Tier (OST) Middleware in HLA Framework for the Distributed Simulation System Environment

Usman Sikander 보안공학연구지원센터 2014 International Journal of Database Theory and Appli Vol.7 No.5

Promising the developers with the facility of distributed collaborative development for complex simulation applications, HLA (High Level Architecture) provides a baseline supporting the reuse of capabilities available in different simulations with a significant reduction in cost and time. Along with improved execution process and reusability, the induction of object oriented model in the implementation design of HLA also enable to take the advantage of latest object oriented features making design, implementation and maintainability easier and at each level of federate development and execution. The important areas to be addressed for design reconsideration consists of data exchange model and HLA communication layer. The data exchange model comprises of federates in federation and between runtime infrastructure and federation. The Federate Object Model (FOM) architecture is not completely object oriented, the induction of Object Service Tier (OST) middleware may offer a degree of FOM agility which is the ability of an application to adjust according to different FOMs (behaviors for Federates). Whereas, in HLA communication layer, customary HLA systems are based upon bidirectional call/callback interactions between federate. Several enhancements and changes anticipated in object oriented communication layer (OOP-COMM) introduced in Object Service Tier (OST) as compared to native procedures, such as communication mechanisms, data encoding, session handling, distributed environment and performance analysis. The motivation behind the use of core object oriented modeling features and proposed Object Service Tier (OST) middleware is the reuse of legacy systems, features which may further enhance the integration of distributed simulation systems and extension types. So, this paper provides a multidimensional analysis of important design aspects of Object Service Tier (OST) middleware in HLA framework and devises some design constructs of Object Service Tier (OST) using object oriented model. This paper is intended to propose object-oriented model providing generalization through the Object Service Tier (OST) middleware in HLA framework for the distributed simulation system environment.

Type Object Class에 의한 Type Object 디자인 패턴의 런타임 클래스 참조문제의 해결

김윤호,Kim, Yun-Ho 한국정보통신학회 2007 한국정보통신학회논문지 Vol.11 No.3

Type Object 디자인 패턴은 하나의 클래스가 수많은 하위 클래스를 갖거나, 그 하위 클래스의 개수를 소프트웨어 개발시에 예측할 수 없는 상황을 해결하고자 제시된 패턴이다. 그러나, 이 패턴은 적용력과 여러 장점에도 불구하고 인스턴스를 생성하는 클래스와 그 인스턴스의 실제적 클래스가 분리되어 있고 또한 객체 레퍼런스에 의해서 서로 연관되어 있으므로 이에 대한 관리를 위한 메커니즘과 패턴의 이해에 있어서 많은 복잡성을 갖는 문제점이 있다. 본 논문에서는 이러한 문제를 해결하기 위한 Type Object Class의 설계와 구현을 제시한다. 즉, Type Object 패턴의 Type Class와 Object Class로부터 Type Object Class를 설정하고, 이를 런타임에 생성되고 사용되게 함으로써, 인스턴스들이 객체 지향프로그래밍 언어에서 제공하는 고유의 클래스를 참조하게 되어 별도의 클래스 참조 메커니즘을 가질 필요가 없도록 하였다. 따라서, 별도의 클래스 참조 메커니즘을 개발하는 부담과 이 메커니즘의 동작으로 인한 실행 상의 성능 저하의 문제가 개선되는 효과가 있다. The Type Object Design Pattern is proposed to provide a solution on the situation in the case that one class has too many subclasses or the number of subclasses are undefined. Although this pattern has many advantages in terms of applicability and dynamic object behavior, it has a weak point in runtime pattern operation that it has to build and maintain a class reference mechanism in runtime to reference the class (de facto 'object') of instances. To solve that problem, this paper addresses the solution of the runtime class reference problem of Type Object Pattern. it defines a new class of Type Object Class (TOC) from Type Class and Object Class in Type Object pattern and presents the methods of creating, compiling, and memory-loading the TOC. It depends on built-in class reference mechanism of object-oriented programming language, and is not necessary to fit with the additional mechanism. Consequently, we need not to set up the additional class reference mechanism and system performance is enhanced due to it.