Convolutional Neural Network to Retrieve Water Depth in Marine Shallow Water Area From Remote Sensing Images

Abstract

The local connection characteristics of convolutional neural network (CNN) are linked with the local spatial correlation of image pixels for water depth retrieval in this article. The method has greater advantages and higher precision than traditional retrieval methods. Traditional remote sensing empirical models require manual extraction of retrieval factors and the process is complex. This article proposes a model based on CNN, which uses different remote sensing images in four spectral bands, red, green, blue, and near-infrared, to retrieve the water depth. In general, CNN is mostly used for image recognition and classification tasks, which can make full use of the local spatial correlation between pixels. The method in this article exploits this feature of CNN for water depth retrieval, taking into consideration the nonlinear relationship between the radiance value and water depth value from adjacent and central pixels. In this article, remote sensing image data, measured water depth data, and lidar sounding data are used as input data to build the model. Then, the retrieval error is analyzed and the parameters are adjusted to further optimize the model. Quantitative analysis and experimental results show that the accuracy of the CNN model in shallow sea areas retrieval is improved by more than 50%. The mean absolute error can reach within 0.8 m. Finally, the model is shown to be highly portable and capable of retrieving water depth data with resolution equal to the spatial resolution of the remote sensing image using only a small amount of input water depth data.

Confidence Measure of the Shallow-Water Bathymetry Map Obtained through the Fusion of Lidar and Multiband Image Data

“…Estimation from R rs . Different from the explicit empirical algorithms shown above, the machine learning approach (MLA, which in this manuscript collectively stands for neural networks, machine learning, and deep learning) is another data-based approach for the estimation of H from remote sensing measurements [29,30,44,45]. Unlike EEA, the algorithm dependence or relationships and coefficients of MLA are hidden in the computer programming architecture (various layers and neurons), so it is not obvious how H imager varies with R rs or spectral radiance.…”

Deriving Highly Accurate Shallow Water Bathymetry From Sentinel-2 and ICESat-2 Datasets by a Multitemporal Stacking Method

“…The widely used linear regression model was adopted to characterize the relationship between Sentinel-2 multispectral bands and ICESat-2 prior elevations. In many previous studies, three visible bands (Blue: Band 2, Green: Band 3, Red: Band 4) and one near-infrared (NIR: Band 8) have been extensively used to construct the linear model for bathymetry inversion [11,13,18,21]. According to a previous study [39], Sentinel-2 coastal aerosol Band 1 exhibited an advantage in satellite-derived bathymetry mapping due to its great water penetration.…”

“…During the past decades, with recent advancements of satellite technologies, satellite-based methods are becoming an applicable way to obtain shallow water bathymetry [6][7][8][9][10][11][12][13][14][15][16][17][18][19], especially at large scales. In previous studies, many empirical models were developed to derive the bathymetry according to the association between the spectral reflectance and depth in shallow water areas [19][20][21].…”

Super-resolution reconstruction of a digital elevation model based on a deep residual network

“…In the urban scene classification based on LiDAR point cloud data, the combination of CNN and RNN can effectively realize the efficient semantic analysis of large-scale 3D point cloud [48] and the combination of Mask R-CNN and LiDAR has great potential for mapping anthropogenic and natural landscape features [49]. In the field of bathymetric survey, deep learning methods are becoming more and more active, for example, some experts and scholars obtained high-resolution water depth data by introducing convolution neural network to process the remote sensing image data and water depth data in shallow water area [50]. It can be seen that neural network model can play an important role in feature extraction of different data, but there is no universal neural network model that could be involved in the feature extraction of various types of data.…”