水利污染遥感监测方法的设计与仿真研究

Design and Simulation of Remote Sensing Monitoring Method for Water Pollution

当前的水利污染遥感监测方法未预处理水利遥感图像,导致污染特征提取结果出现较大误差,污染监测结果不理想。提出一种基于RAISR算法的水利污染遥感监测方法。通过无人机航空遥感系统获取水利遥感图像,采用RAISR算法预处理水利遥感图像,根据训练图像在学习阶段形成滤波器,基于此重建水利污染遥感图像。同时绘制水利遥感图像的直方图,通过最大累间方差法确定阈值,计算获取模型的基本参数,利用EM算法反复迭代直至输出最终参数。通过高斯混合模型计算t-link权重,分割水利遥感图像,提取水利污染遥感特征,实现水利污染的遥感监测。仿真结果表明,所提方法的水利污染监测精度较高,为水利污染的治理提供可靠依据。

The current remote sensing monitoring method of water pollution does not preprocess the water remote sensing image, resulting in large errors in the extraction results of pollution features and unsatisfactory pollution monitoring results. In this regard, a remote sensing monitoring method of water pollution based on RAISR algorithm was proposed. According to the UAV aerial remote sensing system, the water conservancy remote sensing image was obtained. RAISR algorithm was used to preprocess water conservancy remote sensing images. Through the training image, the filter in the learning stage was formed to reconstruct the remote sensing image of water pollution. The histogram of water conservancy remote sensing image was also drawn. The maximum cumulative variance method was introduced to determine the threshold, and the basic parameters of the model were obtained. The EM algorithm was used to iterate until the final parameters were output. According to the Gaussian mixture model, the t-link weight was calculated to segment the water conservancy remote sensing image, extracting the remote sensing characteristics of water conservancy pollution. Eventually, the remote sensing monitoring of water conservancy pollution was realized. The simulation results show that this method has high monitoring accuracy.