Research on Cloud Detection for HY-1C CZI Remote Sensing Images Collected over Lands

Objective The Coast Zone Imager (CZI) onboard the Chinese first marine aqua-color satellite HY-1C started operational operations in June 2019. The large amount of coastal, land, and ocean data it obtained is of great significance for marine disaster and environmental monitoring research. CZI data can be affected by cloud, causing subsequent influences on its applications. How to efficiently identify cloud in remote sensing images is extremely important for the application of CZI images. Most of the existing cloud detection algorithms are based on RGB images or multi-spectral images including thermal infrared band. There is little research on cloud detection algorithms for RGB-NIR four-band remote sensing images, such as HY-1C CZI. The objective of this paper is thus to propose an unsupervised cloud detection method for HY-1C CZI remote sensing images by making full use of NIR band information. Method The method includes four processes: training samples selection, feature extraction, Support Vector Machine (SVM) classification, and post-processing. In the selection of training samples, combined with the dark channel reflectivity, normalized vegetation index and whiteness index of the image, this paper proposes an automatic training sample extraction algorithm, which innovatively uses the whiteness index to obtain detail information, and accurately extract cloud/non-cloud samples through a gradual refinement process. For feature extraction, the spatial spectrum feature information of CZI remote sensing image is selected, including reflectance, spectral index, texture and structure features, to characterize remote sensing image features, and maximize the feature difference between cloud and non-cloud regions. Based on the sample and its feature description, SVM is used to initially classify the CZI remote sensing data, and on this basis, the guided filtering, hole filling and geometric judgment post-processing are performed to obtain the final high-precision cloud detection results. Result This paper applies the algorithm to four typical scenarios (vegetation, soil, wetland, and ice and snow scenarios), and compares and analyzes it with the currently popular unsupervised cloud detection algorithms. The research results found that the cloud detection algorithm proposed in this paper has higher accuracy in these four typical land scenarios in terms of both qualitative or quantitative comparisons.Conclusion The advantage of this algorithm is that it can automatically obtain training samples without manual labeling, and it can also make full use of the near-infrared band information to improve the accuracy of the cloud detection algorithm. Experiments conducted over four typical scenarios rove the effectiveness of the methods for cloud detection from HY-1C CZI data. © 2023 National Remote Sensing Bulletin. All rights reserved.