首页> 外文会议>International Symposium on Intelligent Signal Processing and Communication Systems >A moving object detection method based on level set in dynamic scenes
【24h】

A moving object detection method based on level set in dynamic scenes

机译:一种基于动态场景中级别集的移动物体检测方法

获取原文

摘要

The article is to demonstrate a way to detect moving objects in dynamic scene. This method integrates the means of background difference and level set. First, we adopt background difference, which is to connect the color of video image and movement mode, and then we can renew the dynamic scene by judging the threshold so that the foreground contour initial model can be obtained. Next, the level set can be evolved, through which the more exact foreground objective can be detected. Because of the introduction of movement mode during the course of obtaining the background, the judgment of the background of dynamic scene is more precise. The experiment result suggests that, under the dynamic scene, like ripples and light, the means of level set is the best real-time way to detect the moving objects. However, it is unlikely to detect more than one object in dynamic scene, which is also the future research direction.
机译:文章是展示一种方法来检测动态场景中的移动物体。该方法集成了背景差异和级别集的手段。首先,我们采用背景差异,即要连接视频图像和移动模式的颜色,然后我们可以通过判断阈值来更新动态场景,从而可以获得前景轮廓初始模型。接下来,可以进化水平集,通过该级别可以检测到更精确的前景目标。由于在获得背景过程中引入运动模式,动态场景背景的判断更精确。实验结果表明,在动态场景下,像纹波和光一样,水平集的手段是检测移动物体的最佳实时方法。然而,在动态场景中不太可能检测多个对象,这也是未来的研究方向。

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