首页> 外国专利> DETECTION METHOD OF FORWARD FLIGHT OBJECT USING VISION-BASED INTEREST SEARCH SELECTION ALGORITHM

DETECTION METHOD OF FORWARD FLIGHT OBJECT USING VISION-BASED INTEREST SEARCH SELECTION ALGORITHM

机译:基于视觉的兴趣搜索选择算法的前向飞行物检测方法

摘要

Disclosed is a traffic collision alerting method capable of warning a collision against a bird and a small and light aerial vehicle without a radio communication device through an image, unlike an expensive conventional means, which is dependent on the radio communication and operated by being mounted on and interconnected with a counterpart aerial vehicle. With regard to a method for detecting a flight object, the present invention provides a method for detecting a flight object in front based on vision, which includes a step of detecting an interest region (Step 1); a step of selecting an interest search window in the interest region detected in the step 1 (Step 2); a step of extracting a feature vector in the interest region window selected in the step 2 (Step 3); and a step of detecting the flight object by a classifier using the vector extracted in the step 3 (Step 4).
机译:公开了一种交通冲突警报方法,该方法能够通过图像来警告没有无线电通信设备的鸟类和小型轻型飞行器的碰撞,这不同于昂贵的传统方法,该方法依赖于无线电通信并且通过安装在无线电通信设备上进行操作。并与对应的飞行器互连。关于检测飞行物体的方法,本发明提供了一种基于视觉的检测前方飞行物体的方法,该方法包括检测感兴趣区域的步骤(步骤1);在步骤1中检测出的兴趣区域中选择兴趣搜索窗口的步骤(步骤2);在步骤2中选择的兴趣区域窗口中提取特征向量的步骤(步骤3);以及通过分类器使用在步骤3中提取的矢量来检测飞行物体的步骤(步骤4)。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

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