论文部分内容阅读
雪面上人或动物的活动信息涉及冬季动物迁移追踪、人员活动轨迹检测、安全防卫等各领域应用,利用遥感技术监测人或动物的活动痕迹可以及时掌握动物的迁移情况、江上人员活动范围等信息,进而可为相应的管理决策提供依据。利用高分辨率遥感技术,通过图像滤波等增强处理,可以识别冰雪覆盖区域经人或动物活动后产生的微弱线状扰动痕迹,为后续分析提供帮助。提出了利用高斯差分滤波器(DoG)进行冬季冰雪覆盖区域扰动痕迹识别的监测方法。通过对吉林省龙井市结冰江面上活动痕迹的实验表明,当σ取1.5时,DoG带通滤波频率与人员过江痕迹频率一致,对痕迹增强效果最明显,此时的滤波器为最佳滤波器,痕迹总体提取精度达到83.67%,优于传统的Laplacian算子、Sobel算子和Prewitt算子滤波方法。说明通过DoG滤波器处理,能够有效地突出雪面上人或动物的活动痕迹,可为进一步识别人或动物活动类型、追踪路线以及相关部门的安全巡逻提供服务。
Snow or snow on the human or animal activity information related to the winter animal tracking migration, personnel activity track detection, security and defense applications, the use of remote sensing technology to monitor human or animal traces of activity can be timely grasp of the relocation of animals, Jiang personnel range of activities Information, which in turn can provide the basis for the appropriate management decisions. The use of high resolution remote sensing technology, through image filtering and other enhancement processing, can identify the traces of weak linear disturbances caused by the movement of people and animals in the snow-covered area, which can be used for subsequent analysis. A method of monitoring the disturbance traces in winter and snow covered area by using Gaussian differential filter (DoG) is proposed. Experiments on the traces of activities on the icing river in Longjing City, Jilin Province show that when σ is taken as 1.5, the frequency of DoG band-pass filtering is the same as the frequency of human over-river traces, and the effect of trace enhancement is most obvious. In this case, the filter is the best The overall accuracy of the filter and trace extraction is 83.67%, which is better than the traditional Laplacian, Sobel and Prewitt filter methods. Description DoG filter processing can effectively highlight human or animal traces of snow on the snow, which can be used to further identify the type of human or animal activity, follow-up routes and the security patrols of relevant departments.