• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Inventive Computing and Informatics >Object tracking based hybrid context and image patching models
【24h】

Object tracking based hybrid context and image patching models

机译：基于对象跟踪的混合上下文和图像修补模型

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Object tracking is essential step of many computer vision applications. It deals with locating objects across the frames in videos. There are several useful applications in day-to-day life such as: human-computer interaction, security and surveillance, video, augmented reality, traffic management, aviation control medical imaging etc. There are various challenges when it comes to dealing with object tracking, such as image illumination, motion object, occlusion, scaling etc. In this paper a study of a hybrid model for object tracking is performed. The hybrid model is based two existing algorithms: saliency prior context based model and patching based model. The algorithm generates the similarity based map for the image, then tracks the object in form of patches, whose combined effect is used as a vote map. Then using image signature and spatio-temporal computation a confidence map is generated to predict the next position. Numerous experiment results show that hybrid algorithm achieves significant improvement our saliency prior context based tracker.
机译：对象跟踪是许多计算机视觉应用程序的重要步骤。它处理在视频中的帧中定位对象。在日常生活中有几种有用的应用，如：人机互动，安全和监视，视频，增强现实，交通管理，航空控制医学成像等。在处理对象跟踪方面存在各种挑战（例如图像照明，运动对象，闭塞，缩放等）在本文中，执行用于对象跟踪的混合模型。混合模型是基于两个现有的算法：显着的基于上下文基础的模型和修补程序模型。该算法生成基于相似性的图像的映射，然后以修补程序的形式跟踪对象，其组合效果用作投票映射。然后使用图像签名和时空计算，生成置信度图以预测下一个位置。许多实验结果表明，混合算法实现了显着改善我们的显着性基于上下文的跟踪器。

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