• 한국통신학회논문지
• /
• 제36권7C호
• /
• pp.421-427
• /
• 2011

실시간 객체 인식 및 추적은 컴퓨터 비전 응용 산업이 발달하면서 그 중요성이 더해지고 있다. 객체 추적을 위해 많이 이용되고 있는 알고리즘으로 Mean-Shift 알고리즘이 있다. Mean-Shift 알고리즘을 기반으로 한 객체 추적 알고리즘은 구현이 간단하고, 적은 계산 복잡도를 갖는 장점이 있다. 따라서 실시간 객체 추적 시스템에 적합하다고 할 수 있지만, 지역 모드로의 수렴만을 보장하는 특성으로 인해 객체의 수가 많은 경우 좋은 성능을 나타내지 못하는 단점을 가지고 있다. 그러므로 본 논문에서는 다중 이동 객체를 실시간으로 추적하기 위한 광류기반의 움직임 추정 기법을 제안한다. 제안된 알고리즘의 성능을 확인하기 위해 다중 이동 객체의 인식 실험 결과 유사도는 0.96으로 기존의 Mean-Shift 알고리즘에 비해 약 13.4% 정도 유사도가 개선되었고 평균 픽셀 오류도 3.07로 또한 50% 이상 감소하였다. 향후 알고리즘을 개선하여 처리 속도를 더욱 줄임으로써 매우 빠른 이동 객체 인식과 상황 인지 알고리즘을 추가한다면 보다 효율적인 인식 및 추적 시스템을 구축할 수 있을 것으로 사료된다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 ISIS 2003
• /
• pp.110-113
• /
• 2003

Most traffic accidents are caused by drivers' carelessness and lack of information on the surrounding objects. In this paper we proposed a model of human intention recognition through case-base learning and to build up an experiment system. The system can help us recognize object's intention (e.g. turn left, turn right or straight) by using detected data about human's motion, speed of the car and the distance between the car and the intersection. Furthermore, we included an example using case-base learning in this paper to improve the precision of recognition as well as an example to explain the use of the system. PC can be used to predict the driving reaction beforehand and send a warning signal to the driver in time if there is any danger.

• 한국정보통신학회논문지
• /
• 제13권8호
• /
• pp.1659-1665
• /
• 2009

본 논문에서는 집적영상기술을 이용한 광학적 3D 영상 상관기의 구현 방법을 제안한다. 제안하는 방법에서는 참조 3D 물체와 신호 3D 물체의 요소 영상들이 렌즈 배열을 통하여 얻어지고 이 요소 영상들을 디스플레이 패널에 다시 표시하여 출력 평면에서 광학적으로 고해상도 출력 평면 영상을 얻는다. 이렇게 얻어진 참조 및 신호 물체의 출력 평면 영상에 대하여 상호 상관관계를 수행하여 3D 물체를 인식한다. 제안한 방법은 기존 방법과 비교하여 높은 해상도의 출력 평면 영상을 사용하기 때문에 정확한 3D 물체 인식이 가능하며, 실시간 3D 물체 인식 시스템을 광학적으로 구현할 수 있다. 제안하는 방법의 유용함을 보이기 위하여 광학 실험을 수행하고 그 결과를 보고한다.

• 제어로봇시스템학회논문지
• /
• 제17권3호
• /
• pp.229-235
• /
• 2011

To make a robust object tracking and identifying system for an intelligent robot and/or home system, heterogeneous sensor fusion between visible ray system and infrared ray system is proposed. The proposed system separates the object by combining the ROI (Region of Interest) estimated from two different images based on a heterogeneous sensor that consolidates the ordinary CCD camera and the IR (Infrared) camera. Human's body and face are detected in both images by using different algorithms, such as histogram, optical-flow, skin-color model and Haar model. Also the pose of human body is estimated from the result of body detection in IR image by using PCA algorithm along with AdaBoost algorithm. Then, the results from each detection algorithm are fused to extract the best detection result. To verify the heterogeneous sensor fusion system, few experiments were done in various environments. From the experimental results, the system seems to have good tracking and identification performance regardless of the environmental changes. The application area of the proposed system is not limited to robot or home system but the surveillance system and military system.

• 한국건축시공학회:학술대회논문집
• /
• 한국건축시공학회 2021년도 봄 학술논문 발표대회
• /
• pp.40-41
• /
• 2021

The construction industry has the highest occupational accidents/injuries among all industries. Korean government installed surveillance camera systems at construction sites to reduce occupational accident rates. Construction safety managers are monitoring potential hazards at the sites through surveillance system; however, the human capability of monitoring surveillance system with their own eyes has critical issues. Therefore, this study proposed to build a deep learning-based safety monitoring system that can obtain information on the recognition, location, identification of workers and heavy equipment in the construction sites by applying multiple-object tracking with instance segmentation. To evaluate the system's performance, we utilized the MS COCO and MOT challenge metrics. These results present that it is optimal for efficiently automating monitoring surveillance system task at construction sites.

• 지능정보연구
• /
• 제11권3호
• /
• pp.117-127
• /
• 2005

이동기기를 기반으로 하는 기술이 급격히 발달함에 따라 일반 컴퓨팅 환경의 그래픽이나 영상처리 기술들을 활용한 모바일 컨텐츠 개발이 증가하고 있다. 본 논문에서는 상호 연동성을 지원하는 국내 표준 무선인터넷 플랫폼인 WIPI를 기반으로 하여 영상처리와 얼굴인식 기술을 응용한 얼굴탐지 및 인식 지원 미들웨어를 설계하였다. 생성된 얼굴인식 미들웨어는 휴대폰 카메라를 이용한 인식보안 및 다양한 컨텐츠로 활용될 수 있도록 객체지향 개념을 도입해 각각의 미들웨어가 얼굴인식 미들웨어를 공유할 수 있다. 이는 컨텐츠 설계에 있어 주요 프로세스를 분리함으로써 컨텐츠의 개발 기간 및 비용을 단축시킬 수 있으며, 컨텐츠 보호 및 타 기업과의 기술이전에 적용될 수 있다. 얼굴인식 미들웨어는 서비스에 따라 얼굴탐지 모듈과 얼굴인식 모듈로 구성되며, WIPI 플랫폼 상에서의 얼굴인식 미들웨어 응용 컨텐츠 설계 방법에 대하여 제시한다.

• 전기학회논문지
• /
• 제66권4호
• /
• pp.672-681
• /
• 2017

This paper is concerned with the design methodology of face recognition system based on pose estimation. In 2-dimensional face recognition, the variations of facial pose cause the deterioration of recognition performance because object recognition is carried out by using brightness of each pixel on image. To alleviate such problem, the proposed face recognition system deals with Learning Vector Quantizatioin(LVQ) or K-Nearest Neighbor(K-NN) to estimate facial pose on image and then the images obtained from LVQ or K-NN are used as the inputs of networks such as Convolution Neural Networks(CNNs) and Radial Basis Function Neural Networks(RBFNNs). The effectiveness and efficiency of the post estimation using LVQ and K-NN as well as face recognition rate using CNNs and RBFNNs are discussed through experiments carried out by using ICPR and CMU PIE databases.

• 한국자동차공학회논문집
• /
• 제24권2호
• /
• pp.127-136
• /
• 2016

The camera based object detection systems should satisfy the recognition performance as well as real-time constraints. Particularly, in safety-critical systems such as Autonomous Emergency Braking (AEB), the real-time constraints significantly affects the system performance. Recently, multi-core processors and system-on-chip technologies are widely used to accelerate the object detection algorithm by distributing computational loads. However, due to the advanced hardware, the complexity of system architecture is increased even though additional hardwares improve the real-time performance. The increased complexity also cause difficulty in migration of existing algorithms and development of new algorithms. In this paper, to improve real-time performance and design complexity, a task scheduling strategy is proposed for visual object tracking systems. The real-time performance of the vision algorithm is increased by applying pipelining to task scheduling in a multi-core processor. Finally, the proposed task scheduling algorithm is applied to crosswalk detection and tracking system to prove the effectiveness of the proposed strategy.

• 한국컴퓨터정보학회논문지
• /
• 제23권1호
• /
• pp.41-48
• /
• 2018

Recently, surveillance systems are widely used, and one of the key technologies in this surveillance system is to recognize and track objects. In order to track a moving object robustly and efficiently in a complex environment, it is necessary to extract the feature points in the interesting object and to track the object using the feature points. In this paper, we propose a method to track interesting objects in real time by eliminating unnecessary information from objects, generating feature point descriptors using only key feature points, and reducing computational complexity for object recognition. Experimental results show that the proposed method is faster and more robust than conventional methods, and can accurately track objects in various environments.

• Agricultural and Biosystems Engineering
• /
• 제2권2호
• /
• pp.69-74
• /
• 2001

Detection of operation lines in farm works, object recognition and obstacle avoidance are essential pre-requisite technologies for unmanned agricultural machinery. A CCD camera, which has been largely used for these functions, is expensive and has difficulty in real-time signal processing. In this study, a laser range sensor was selected as the guiding vision for unmanned agricultural machinery such as a tractor. To achieve this capability, algorithms for distance measurement, signal filtering, object recognition, and obstacle avoidance were developed. Computer simulations were carried out to evaluate performance of the algorithms. Experiments were also conducted with various materials and shapes, Laser beam lost its intensity for poor reflective materials, resulting in less range value than actual, so a compensation technique was considered to be necessary. Object detection system was fabricated on an agricultural tractor and the performance was evaluated. As test result for obstacle detection and avoidance in field, to detect and avoid obstacle for path finding with guiding system for unmanned agricultural machinery was enable.