| 摘要: |
| 潮沟是潮滩的主要地貌类型之一, 潮沟信息的检测对湿地生态环境的监测和保护有着重要意义。本文以长江口九段沙下沙北部发育比较复杂的潮沟为研究对象, 使用2016年1月26日Landsat8全色波段数据, 采用了一种融合小波变换、最大类间方差法和数学形态学的方法对潮沟边缘信息进行了检测。小波变换中通过减小高层小波系数, 达到弱化潮滩灰度变化的目的; 通过增大低层小波系数,达到增强潮沟信息的目的; 利用数学形态学和图像之间的几何(加减乘)运算, 完成潮沟的提取; 最后,分别在原始数据和检测结果相应的位置取样进行精度验证, 样方潮沟面积一致性精度平均为92.1%。 |
| 关键词: 潮沟 小波变换 最大类间方差法 数学形态学 基于连通域的断点连接 |
| DOI：10.11759/hykx20170412002 |
| 分类号: |
| 基金项目:国土资源部公益性行业科研专项(201211009) |
|
| Research on tidal channel detection based on wavelet transformation and the mathematical morphology method |
|
HE Si-hai,HAN Zhen,ZHU Yan-jiang,HU Xu-ran,Ding Ru-yi
|
| Abstract: |
| Tidal channels are one of the major landforms in tidal flats, and their detection plays a key role in monitoring and protecting the wetland ecological environments. Considering a complex tidal channel from northern Xiasha in the Jiuduan Sha of the Yangtze River estuary as an object, this paper adopted a fusion of wavelet transformation, the Otsu method, and the mathematical morphology method to detect the edge information using the Landsat8 panchromatic band data from January 26, 2016. During the process of wavelet transformation, reducing the high-level detail coefficients resulted in weakening the tidal flat information, whereas increasing the low-layer detail coefficients resulted in enhancing the tidal channel information. The mathematical morphology method and the operations, such as addition, subtraction, and multiplication of images, were used to complete the extraction of the tidal channel. The accuracy of the tidal flat area consistently averaged to 92.1%, which was verified by sampling the original data and the corresponding results. |
| Key words: tidal flat wavelet transform Otsu method mathematical morphology method verify the accuracy breakpoint connection based on connected domain |
