论文部分内容阅读
目前基于目视解释或光谱分类的养殖信息提取效率低,难以克服由于地物混杂带来的“椒盐”噪声现象且难以融合地学知识。针对养殖信息提取中存在的问题,首先在分析现有养殖信息提取方法和案例推理CBR(Case-Based Reasoning)用于遥感图像处理的基础上,提出基于遥感案例推理的海岸带养殖信息提取的研究思路;其次,结合养殖区域的空间特征和属性特征,构建案例的表达模型以及CBR相似性推理模型;最后,对不属于案例构建区的粤西沙田镇进行养殖信息提取的CBR实验,精度达到84.56%。对比CBR方法和传统监督分类方法可知,CBR方法是实现海岸带养殖信息快速准确提取的一种有效手段。
At present, the extraction efficiency of aquaculture information based on visual interpretation or spectral classification is low, so it is difficult to overcome the phenomenon of “salt and pepper” noise caused by the mixed features and difficult to integrate knowledge. In order to solve the existing problems in aquaculture information extraction, firstly, based on the analysis of the existing farming information extraction methods and Case-Based Reasoning (CBR) for remote sensing image processing, this paper proposes a case-based reasoning-based coastal aquaculture information extraction Secondly, CBR model of CBR was constructed based on the spatial and attribute features of farmed area. Finally, the CBR experiment of extracting aquaculture information was carried out in Shatian Town, western Guangdong, which is not a case-building area, with a precision of 84.56 %. Compared with the CBR method and the traditional supervised classification method, the CBR method is an effective method to realize the rapid and accurate extraction of coastal zone breeding information.