• 융합신호처리학회논문지
• /
• 제15권4호
• /
• pp.112-118
• /
• 2014

본 연구는 칼라기반에서 단일 이동객체 추적을 다루고 있다. 우선 매 영상에서 이동객체 영상의 밝기 변화에 따른 추적 약점을 개선하기 위해 기존의 Camshift 알고리즘을 보완하였다. 보완된 알고리즘도 추적중인 물체와 색상이 같은 주변 물체가 존재할 경우 불안정한 추적을 보여주었는데 본 연구에서는 이를 해결하기 위해 Optical Flow기반의 KLT 알고리즘과 병합하는 방법을 제시하였다. 픽셀기반의 특징점 추적을 수행하는 KLT 알고리즘은 칼라기반의 Camshift의 단점을 보완할 수 있다. 실험 결과 제안된 병합 방법은 기존의 추적단점을 보완하였으며 추적성능이 개선됨을 실험으로 확인하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권3호
• /
• pp.1121-1141
• /
• 2020

We propose a robust visual object tracking algorithm fusing a convolutional neural network tracker trained offline from a large number of video repositories and a color histogram based tracker to track objects for mixing immersive audio. Our algorithm addresses the problem of occlusion and large movements of the CNN based GOTURN generic object tracker. The key idea is the offline training of a binary classifier with the color histogram similarity values estimated via both trackers used in this method to opt appropriate tracker for target tracking and update both trackers with the predicted bounding box position of the target to continue tracking. Furthermore, a histogram similarity constraint is applied before updating the trackers to maximize the tracking accuracy. Finally, we compute the depth(z) of the target object by one of the prominent unsupervised monocular depth estimation algorithms to ensure the necessary 3D position of the tracked object to mix the immersive audio into that object. Our proposed algorithm demonstrates about 2% improved accuracy over the outperforming GOTURN algorithm in the existing VOT2014 tracking benchmark. Additionally, our tracker also works well to track multiple objects utilizing the concept of single object tracker but no demonstrations on any MOT benchmark.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2007년도 추계종합학술대회
• /
• pp.277-280
• /
• 2007

특정컬러정보 검출기반의 이동객체 탐색 알고리듬을 구현한다. 입력 이미지에 대해 조도변화 및 노이즈 제거 등을 위해 전처리 과정을 거치고, 이동객체 탐색은 R,G,B 각 채널의 영상차를 이용하여 객체를 검색한다. 실험 결과 검색 속도는 윤관선 탐색 및 정합법에 비해 15% 향상되었고 안정적이다. 또한 컬러 정보 기반의 객체 탐색이 가능함을 제시하였다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2006년 학술대회 논문집 정보 및 제어부문
• /
• pp.434-436
• /
• 2006

In a visual driver-assistance system, separating moving objects from fixed objects are an important problem to maintain multiple hypothesis for the state. Color and edge-based tracker can often be "distracted" causing them to track the wrong object. Many researchers have dealt with this problem by using multiple features, as it is unlikely that all will be distracted at the same time. In this paper, we improve the accuracy and robustness of real-time tracking by combining a color histogram feature with a brightness of Optical Flow-based feature under a Sequential Monte Carlo framework. And it is also excepted from Tracking as time goes on, reducing density by Adaptive Particles Number in case of the fixed object. This new framework makes two main contributions. The one is about the prediction framework which separating moving objects from fixed objects and the other is about measurement framework to get a information from the visual data under a partial occlusion.

• 대한전자공학회논문지SP
• /
• 제45권5호
• /
• pp.63-71
• /
• 2008

본 논문에서는 다관절체 추적을 위해 기존에 물체 추적에 자주 이용되는 파티클 필터를 확장한 계층적 파티클 필터 방법을 제안한다. 칼라 특징은 부분 겹침, 회전등에 강건한 특징을 가지고 있어서, 칼라 기반 파티클 필터는 물체 추적에 널리 쓰이고 있다. 다관절체 추적에서 상태 벡터는 높은 차원을 가지기 때문에 기존의 파티클 필터를 이용해 바람직한 추적 결과를 얻기 위해서는 많은 수의 샘플이 요구된다. 이러한 문제점을 해결하기 위해, 본 논문에서는 이미 알고 있는 다른 신체 부위의 위치를 이용해 계층적으로 신체 부위를 추적한다. 계층적 추적 방법에 의해 복잡한 환경에서 강건한 추적을 위한 샘플의 수를 줄일 수 있었다. 또한 포즈를 인식하기 위해 상박과 하박의 각도를 이용한 SVM(Support Vector Machine)을 이용해 8개의 포즈를 분류한다. 실험 결과는 세안한 방법이 기존의 칼라 기반의 파티클 필터보다 효율적임을 보여준다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2016년도 춘계학술발표대회
• /
• pp.743-744
• /
• 2016

Object tracking in video sequences is a challenging task and has various applications. Particle filtering has been proven very successful for non-Gaussian and non-linear estimation problems. In this study, we first try to develop a color-based particle filter. In this approach, the color distributions of video frames are integrated into particle filtering. Color distributions are applied because of their robustness and computational efficiency. The model of the particle filter is defined by the color information of the tracked object. The model is compared with the current hypotheses of the particle filter using the Bhattacharyya coefficient. The proposed tracking method directly incorporates the scale and motion changes of the objects. Experimental results have been presented to show the effectiveness of our proposed system.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2018년도 추계학술발표대회
• /
• pp.808-810
• /
• 2018

Pedestrian tracking is a particular problem of object tracking, and an important component in various vision-based applications, such as autonomous cars or surveillance systems. After several years of development, pedestrian tracking in videos is still a challenging problem because of various visual properties of objects and surrounding environment. In this research, we propose a tracking-by-detection system for pedestrian tracking, which incorporates Convolutional Neural Network (CNN) and color information. Pedestrians in video frames are localized by a CNN, then detected pedestrians are assigned to their corresponding tracklets based on similarities in color distributions. The experimental results show that our system was able to overcome various difficulties to produce highly accurate tracking results.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2005년도 추계종합학술대회
• /
• pp.347-350
• /
• 2005

This paper proposes a new tracking algorithm using the Kalman estimator based color SSD algorithm. The Kalman estimator includes the color information as well as the position and size of the face region in its state vector, to take care of the variation of skin color while faces are moving. Based on the estimated face position, the color SSD algorithm finds the face matching with the one in the previous frame even when the color and size of the face region vary. The features of a face region extracted by the color SSD algorithm are used to update the state of the Kalman estimator.

• 제어로봇시스템학회논문지
• /
• 제18권6호
• /
• pp.562-570
• /
• 2012

This paper proposes a robust object tracking algorithm using object features and on-line learning based particle filter for mobile robots. Mobile robots with a side-view camera have problems as camera jitter, illumination change, object shape variation and occlusion in variety environments. In order to overcome these problems, color histogram and HOG descriptor are fused for efficient representation of an object. Particle filter is used for robust object tracking with on-line learning method IPCA in non-linear environment. The validity of the proposed algorithm is revealed via experiments with DBs acquired in variety environment. The experiments show that the accuracy performance of particle filter using combined color and shape information associated with online learning (92.4 %) is more robust than that of particle filter using only color information (71.1 %) or particle filter using shape and color information without on-line learning (90.3 %).

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 추계종합학술대회 논문집
• /
• pp.1003-1006
• /
• 1999

Hand segmentation and tracking is essential to the development of a hand gesture recognition system. This research features segementation and tracking of hand regions based the hue component of color. We propose a method that employs HSI color model, and segments and tracks hand regions using the hue component of color alone. In order to track the segmented hand regions, we only apply Kalman filter to a region of interest represented by a rectangle region. Initial experimental results show that the system accurately segments and tracks hand regions although it only uses the hue compoent of color. The system yields near real time throghput of 8 frames per second on a Pentium II 233MHz PC.