Bathymetry of the Coral Reefs of Weizhou Island Based on Multispectral Satellite Images

作者：Rongyong Huang, Kefu Yu, Yinghui Wang, Jikun Wang, Mu Lin, Wenhuan Wang · 发表于：Remote Sensing · 年份：2017 · DOI：10.3390/rs9070750 · 被引用次数：45 · 研究领域：Remote Sensing and LiDAR Applications、Remote-Sensing Image Classification、Remote Sensing and Land Use

Shallow water depth measurements using multispectral images are crucial for marine surveying and mapping. At present, relevant studies either depend on the use of other auxiliary data (such as field water depths or water column data) or contain too many unknown variables, thus making these studies suitable only for images that contain enough visible wavebands. To solve this problem, a Quasi-Analytical Algorithm (QAA) approach is proposed in this paper for estimating the water depths around Weizhou Island by developing a QAA to estimate the diffuse attenuation coefficients and simplifying the parameterization of the bathymetric model. The approach contains an initialization sub-approach and a novel global adjustment sub-approach. It is not only independent of other auxiliary data but also greatly reduces the number of unknowns. Experimental results finally demonstrated that the Root Mean Square Errors (RMSEs) were 1.01 m and 0.77 m for the ZY-3 image and the WorldView-3 (WV-3) image, respectively, so the approach is competitive to other QAA bathymetric methods. Besides, the global adjustment sub-approach was also seen to be superior to common smoothing filters: if the Signal-to-Noise Ratio (SNR) is as low as 42, i.e., ZY-3, it can smooth the water depths and improve the accuracies, otherwise can avoid the over-smoothing of water depths.