• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Transactions of the Institute of Measurement and Control >Multiple stationary-non-stationary object-tracking approach in real-time applications through an extendable RGB modelling framework
【24h】

Multiple stationary-non-stationary object-tracking approach in real-time applications through an extendable RGB modelling framework

机译：通过可扩展的RGB建模框架在实时应用中使用多种固定-非平稳对象跟踪方法

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

The main contribution of the present research arises from developing standard object tracking approaches by considering a number of available research works in the area of multiple-object tracking approaches in real-time applications. An automatic robust algorithm is proposed in this investigation, which is able to track multiple synchronized stationary and/or non-stationary objects in high performance quality. The proposed approach is realized based upon an efficient extendable RGB (red/green/blue) modelling framework. This means that the investigated results are easily applicable to tracking the desirable number of objects to be chosen, i.e. the approach is flexible enough to reorganize to track the stationaryon-stationary objects in all frames of the video sequences. This proposed idea unique in that we plan to track all the chosen objects, in a synchronized manner, without a lengthy processing time. In fact, the proposed approach is realized in association with a desirable number of sub-systems, each of them needing to be enabled in parallel. The investigated results indicate that the present approach is now applicable in real-time applications by running a robust estimator to predict the new position of the tracked objects, since the objects are in the presence of long-term environmental problems - situations that are usually too difficult to deal with through standard object tracking algorithms.
机译：本研究的主要贡献来自通过考虑实时应用中多对象跟踪方法领域中的许多可用研究工作而开发的标准对象跟踪方法。在这项研究中提出了一种自动鲁棒算法，该算法能够以高性能质量跟踪多个同步的固定和/或非固定对象。所提出的方法是基于有效的可扩展RGB（红色/绿色/蓝色）建模框架实现的。这意味着所研究的结果很容易适用于跟踪希望选择的对象数量，即该方法足够灵活以进行重组以跟踪视频序列所有帧中的静止/非静止对象。这个提议的想法非常独特，因为我们计划以同步的方式跟踪所有选定的对象，而无需花费很长时间。实际上，所提出的方法是与所需数目的子系统相关联地实现的，每个子系统都需要并行启用。研究结果表明，通过运行鲁棒的估计器来预测被跟踪物体的新位置，本方法现在可用于实时应用，因为这些物体存在长期的环境问题-通常情况也是如此通过标准的对象跟踪算法很难处理。

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