首页> 外国专利> Methods for automated real-time acquisition of marine mammals (acquisition)

Methods for automated real-time acquisition of marine mammals (acquisition)

机译:自动实时采集海洋哺乳动物的方法(采集)

摘要

To monitor the presence of marine mammals based on detecting thermal signature of (TSMM) is manually by wardens far , or has been performed in a simple partial prototype automated only . With such a method , there is a problem in image processing with a quick and reliable combined with ease of use of the altitude. The method according to the invention , in fragments in the image pre- processing step (FPP), image correction and complete image segmentation , which has not been weighted (SEG); in the detection process (DET), based on the detected local contrast change use of threshold algorithm ; in the classification process (CLA), for the subdivision of the two classes , use of the help monitored learning of support vector method with a hyper plane (HE) (SVM); localization of in step (LOC), the distance of the detected thermal distinguishing characteristics (SMM), and automatically calculate the changes that over time and space for the vessel (RV); in the verification process (VER), the moment of decision by the operator specific check of possibility ; and , in the documentation process (DOC), including automatically providing a user-specific image data of the detected thermal signatures of marine mammals (TSMM) (IRV). Infrared camera system with a complete or partial peripheral monitoring (IRC), and an active gyroscope stabilization against the horizon (HZ), are used to produce the data .
机译:通过检测(TSMM)的热特征来监视海洋哺乳动物的存在是由病房手动完成的,或者仅在自动化的简单局部原型中执行。通过这种方法,在图像处理中存在与快速且可靠结合高度的易于使用相结合的问题。根据本发明的方法,在图像预处理步骤(FPP),图像校正和完整图像分割中的片段中,将其加权(SEG);在检测过程(DET)中,基于检测到的局部对比度变化使用阈值算法;在分类过程(CLA)中,对于这两个类别的细分,使用带有超平面(HE)(SVM)的帮助监控学习支持向量方法;在步骤(LOC)中定位所检测到的热区别特征(SMM)的距离,并自动计算随时间和空间变化的容器变化(RV);在验证过程(VER)中,决定时刻由操作者具体检查可能性;并且,在记录过程(DOC)中,包括自动提供检测到的海洋哺乳动物热信号(TSMM)(IRV)的用户特定图像数据。具有完整或部分外围监视(IRC)的主动红外陀螺仪系统以及针对水平线的主动陀螺仪稳定系统(HZ)用于生成数据。

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