INFORMATION TECHNOLOGY

New progress has been made in the research of optical shallow water depth remote sensing inversion


The team of Tang Shilin, researcher of the State Key Laboratory of Tropical Marine Environment, South China Sea Institute of Oceanology, Chinese Academy of Sciences, together with Professor Deng Xiaoxiu of Sun Yat-sen University and other researchers, have made new progress in the research of optical shallow water depth remote sensing inversion, and realized ultra-high spatial resolution (meter-level) remote sensing bathymetry in shallow coastal areas without on-site observation data. Recently, the relevant results were published in IEEE Transactions on Geoscience and Remote Sensing.

Map of 16 optical shallow water study areas worldwide. Photo courtesy of the research team

Underwater terrain exploration is an important basic work for human development and utilization of the ocean. The previous research of Liu Yongming et al. preliminarily constructed a downscale water depth remote sensing inversion model (DBMA), which can realize the remote sensing inversion of optical shallow water depth information without the assistance of measured water depth data, and the model has been verified in some optical shallow water areas in the South China Sea. However, due to the high cost of ultra-high spatial resolution satellite image data, the unclear maximum detection depth of the water bathymetry optimization inversion model, and the error transmission problem of the coupling model, the popularization and application of the DBMA model on the global scale of terrain information extraction is constrained.

In response to the above problems, the researchers further carried out research. Based on the image data of Landsat 8, they proposed a theoretical maximum detection depth calculation method for the water depth optimization inversion model to obtain the theoretical maximum optical detection depth, and verified that the root mean square error of the theoretical maximum detection depth was 3.30 m. Based on the measured data and the results of the maximum detection depth, the correction parameters of the coupling model error transfer are obtained, and the results show that this set of parameters can significantly improve the accuracy of water depth inversion.

Finally, based on the maximum detection depth constraint and correction coefficient, the researchers constructed DBMA-RGB using ultra-high spatial resolution RGB (Red-green-blue) images provided by Landsat 8 and Google Earth, and compared and verified it in 16 different optical shallow water regions around the world, and the results showed that the accuracy value of water depth inversion was similar to the statistical model of spaceborne lidar calibration.

Tang Shilin, the corresponding author of the paper, said that the research will help develop global optical shallow water ultra-high spatial resolution water depth products.

The above research was supported by the National Natural Science Foundation of China, the Hainan Science and Technology Special Project and the Guangzhou Municipal Foundation. (Source: China Science News, Zhu Hanbin, Fu Tian)

Related paper information:https://doi.org/10.1109/TGRS.2023.3315316



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button