• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Mathematical methods for information science and economics >Robust Video Object Detection and Tracking Techniques
【24h】

Robust Video Object Detection and Tracking Techniques

机译：强大的视频对象检测和跟踪技术

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Video object detection and tracking represents an important computer vision domain that has been vividly researched in the last decades. It has promising applications in numerous important fields, such as video compression, video surveillance, human-computer interaction, video indexing and retrieval, medical imaging, traffic monitoring, augmented reality and robotics. Obviously, it consists of two closely related processes. The first one, video object detection involves locating an image object in theframes of a video sequence, while the second one, video tracking, represents the process ofmonitoring the video object spatial and temporal changes during the moviesequence, including its presence, position, size and shape.While an object detection algorithm identifies image objects in video frames, an object tracking procedure must solve the temporal correspondence problem that is the task of matching the target object in successive frames. Numerous video detection and tracking technologies have been developed inrecent years. Object detection can be performed through various approaches, such asxegion-based image segmentation, background subtraction, temporaldifferencing, active contour modelsand the generalized Hough transforms.Video tracking techniques are based on Kalman filtering, Hidden Markov Models, optical flow, template matching, mean-shift trackingand contour tracking.Objecttracking is often a time consuming process due to the amount of data contained by video streams. Also, video tracking represents a difficult process, becausevarious factors such as abrupt object motion, object occlusions or camera motion. There are various types of tracking, depending on the target object character (static or moving) and the camera (fixed or moving). We approached the object detection and tracking domain in our previous works, developing some robust detection and tracking techniques for both static camera and moving camera videos. Thus, we proposed several automatic temporal-differencing based moving object detection approaches for fixed camera video sequences. The object tracking was performed using template matching and various object featuring methods. We used HOG-based, normalized cross-correlation based and 2D Gabor filtering based features for this purpose. Also, we considered video tracking approaches which are able to track successfully both the static and moving objects, in both static-camera and moving camera videos. Thus, we developed a novel semiautomatic object tracking technique based on an improved N-Step Search algorithm and a HOG-based feature extraction. Human detection and tracking, representing an important sub-domain of object detection and tracking, is also widely approached in our research.
机译：视频对象检测和跟踪代表了重要的计算机视觉领域，最近几十年来对此进行了生动的研究。它在视频压缩，视频监控，人机交互，视频索引和检索，医学成像，交通监控，增强现实和机器人技术等众多重要领域中具有广阔的应用前景。显然，它由两个密切相关的过程组成。第一个视频对象检测涉及在视频序列的帧中定位图像对象，而第二个视频跟踪表示在电影序列期间监视视频对象的时空变化的过程，包括其存在，位置，大小和大小。虽然对象检测算法可以识别视频帧中的图像对象，但是对象跟踪过程必须解决时间对应问题，这是在连续帧中匹配目标对象的任务。最近几年开发了许多视频检测和跟踪技术。可以通过各种方法执行目标检测，例如基于像素的图像分割，背景扣除，时间差分，活动轮廓模型和广义的Hough变换。视频跟踪技术基于卡尔曼滤波，隐马尔可夫模型，光流，模板匹配，均值移位跟踪和轮廓跟踪。由于视频流包含的数据量大，对象跟踪通常是一个耗时的过程。而且，视频跟踪代表了一个艰巨的过程，因为各种因素（例如突然的物体运动，物体遮挡或摄像机运动）。跟踪的类型多种多样，具体取决于目标对象的角色（静态或移动）和摄像机（固定或移动）。我们在之前的工作中接触了对象检测和跟踪领域，为静态摄像机和运动摄像机视频开发了一些可靠的检测和跟踪技术。因此，我们提出了几种针对固定摄像机视频序列的基于自动时差的运动对象检测方法。使用模板匹配和各种对象特征化方法执行对象跟踪。为此，我们使用了基于HOG，归一化互相关和基于2D Gabor滤波的功能。此外，我们考虑了视频跟踪方法，该方法能够成功跟踪静态摄像机和移动摄像机视频中的静态和动态对象。因此，我们基于改进的N步搜索算法和基于HOG的特征提取开发了一种新颖的半自动对象跟踪技术。人体检测和跟踪代表了对象检测和跟踪的一个重要子领域，在我们的研究中也被广泛采用。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号