Cloud detection for HY-1C/COCTS over the ocean based on spectral-and-textural-information-guided deep neural network

Abstract

Given that clouds can absorb and scatter radiation signals in the visible and infrared bands, cloud detection is a key preprocessing step for ocean color and sea surface temperature retrievals. In this research, a Spectral-and-Textural-Information-Guided deep neural Network (STIGNet) is designed for cloud detection in global ocean data from the Haiyang-1C (HY-1C)/Chinese Ocean Color and Temperature Scanner (COCTS). Considering the spectral and textural properties of clouds, the model incorporates HY-1C/COCTS spectral data, differences in brightness temperature (BT), local statistical characteristics of BT, and geographical location information–all of which are closely related to cloud features. Notably, an edge learning module is implemented to emphasize edge features during the training process. We construct a HY-1C/COCTS cloud detection dataset to train and test the cloud detection model. In the dataset, labels are generated by combining the Bayesian cloud detection method with a manual mask. Analysis of the resulting cloud detection images indicates that STIGNet exhibits accurate performance across various types of clouds while showing minimal overestimated errors in areas such as ocean fronts or sun glints, where they tend to occur frequently. The ablation experiments performed on physical-based input features and edge learning modules show enhancements in cloud detection accuracy. Evaluation results demonstrate an overall accuracy of 96.64%, with a cloud overestimated error of 1.61% and a cloud missed error of 1.76%. These findings highlight the effectiveness of STIGNet in generating precise cloud masks for HY-1C/COCTS data.

References

[1] AckermanSA, StrabalaKI, MenzelWP, FreyRA, MoellerCC, GumleyLE. Discriminating clear sky from clouds with MODIS. J Geophys Res-Atmos, 1998, 103(D24): 32141-32157

[2] AliA, de BieC, SkidmoreA. Detecting long-duration cloud contamination in hyper-temporal NDVI imagery. Int J Appl Earth Obs Geoinf, 2013, 24: 22-31

[3] BulginCE, MittazJPD, EmburyO, EastwoodS, MerchantCJ. Bayesian cloud detection for 37 years of Advanced Very High Resolution Radiometer (AVHRR) Global Area Coverage (GAC) data. Remote Sens, 2018, 10

[4] ChaiD, NewsamS, ZhangHKK, QiuY, HuangJF. Cloud and cloud shadow detection in Landsat imagery based on deep convolutional neural networks. Remote Sens Environ, 2019, 225: 307-316

[5] Chen LCE, Zhu YK, Papandreou G, Schroff F, Adam H (2018) Encoder-decoder with atrous separable convolution for semantic image segmentation. In: 15th European Conference on Computer Vision (ECCV), Munich, pp 801–818. https://doi.org/10.1007/978-3-030-01234-2_49

[6] ESA (2022) Cloud identification. https://sentinel.esa.int/web/sentinel/technical-guides/sentinel-3-slstr/level-1/cloud-identification. Accessed 22 Jan 2024

[7] Glorot X, Bordes A, Bengio Y (2011) Deep sparse rectifier neural networks. In: Proceedings of the 14th International Conference on Artificial Intelligence and Statistics, Fort Lauderdale, pp 315–323

[8] GuptaR, NandaSJ, ShuklaUP. Cloud detection in satellite images using multi-objective social spider optimization. Appl Soft Comput, 2019, 79: 203-226

[9] Ioffe S, Szegedy C (2015) Batch normalization: accelerating deep network training by reducing internal covariate shift. In: ICML’15: Proceedings of the 32nd International Conference on Machine Learning, Lille, pp 448–456. https://dl.acm.org/doi/10.5555/3045118.3045167

[10] Irish RR, Barker JL, Goward SN, Arvidson T (2006) Characterization of the landsat-7 ETM+ automated cloud-cover assessment (ACCA) algorithm. Am Soc Photogramm Remote Sens 72:1179–1188. https://doi.org/10.14358/PERS.72.10.1179

[11] IshidaH, OishiY, MoritaK, MoriwakiK, NakajimaTY. Development of a support vector machine based cloud detection method for MODIS with the adjustability to various conditions. Remote Sens Environ, 2018, 205: 390-407

[12] JeppesenJH, JacobsenRH, InceogluF, ToftegaardTS. A cloud detection algorithm for satellite imagery based on deep learning. Remote Sens Environ, 2019, 229: 247-259

[13] JoshiPP, WynneRH, ThomasVA. Cloud detection algorithm using SVM with SWIR2 and tasseled cap applied to Landsat 8. Int J Appl Earth Obs Geoinf, 2019, 82

[14] KanuS, KhojaR, LalS, RaghavendraBS, AshaCS. CloudX-net: a robust encoder-decoder architecture for cloud detection from satellite remote sensing images. Remote Sens Appl Soc Environ, 2020, 20

[15] KilpatrickKA, PodestáG, WilliamsE, WalshS, MinnettPJ. Alternating decision trees for cloud masking in MODIS and VIIRS NASA sea surface temperature products. J Atmos Ocean Technol, 2019, 36(3): 387-407

[16] KingMD, PlatnickS, MenzelWP, AckermanSA, HubanksPA. Spatial and temporal distribution of clouds observed by MODIS onboard the terra and aqua satellites. IEEE Trans Geosci Remote Sens, 2013, 51(7): 3826-3852

[17] Kingma D, Ba J (2014) Adam: a method for stochastic optimization. In: 3rd International Conference for Learning Representations, San Diego, pp 1–13

[18] LiWB, SunKM, DuZT, HuXQ, LiWZ, WeiJJ, et al. . PCNet: cloud detection in FY-3D true-color imagery using multi-scale pyramid contextual information. Remote Sens, 2021, 13(18)

[19] LiX, YangXF, LiXT, LuSJ, YeYM, BanYF. GCDB-UNet: a novel robust cloud detection approach for remote sensing images. Knowledge-Based Syst, 2022, 238

[20] LiYS, ChenW, ZhangYJ, TaoC, XiaoR, TanYH. Accurate cloud detection in high-resolution remote sensing imagery by weakly supervised deep learning. Remote Sens Environ, 2020, 250

[21] LiZW, ShenHF, ChengQ, LiuYH, HeYou SC, ZY. Deep learning based cloud detection for medium and high resolution remote sensing images of different sensors. ISPRS-J Photogramm Remote Sens, 2019, 150: 197-212

[22] LinCH, LinBY, LeeKY, ChenYC. Radiometric normalization and cloud detection of optical satellite images using invariant pixels. ISPRS-J Photogramm Remote Sens, 2015, 106: 107-117

[23] LinTY, GoyalP, GirshickR, HeKM, DollarP. Focal loss for dense object detection. IEEE Trans Pattern Anal Mach Intell, 2020, 42(2): 318-327

[24] LiuM, GuanL, LiuJ, LiuJQ, SongQJ, MaCF, et al. . First assessment of HY-1C COCTS thermal infrared calibration using MetOp-B IASI. Remote Sens, 2021, 13(4)

[25] LiuM, MerchantCJ, EmburyO, LiuJQ, SongQJ, GuanL. Retrieval of sea surface temperature from HY-1B COCTS. IEEE Trans Geosci Remote Sens, 2022, 60: 1-13

[26] LuoC, FengSS, LiXT, YeYM, ZhangBQ, ChenZH, et al. . ECDNet: a bilateral lightweight cloud detection network for remote sensing images. Pattern Recognit, 2022, 129

[27] MerchantCJ, EmburyO, RaynerNA, BerryDI, CorlettGK, LeanK, et al. . A 20 year independent record of sea surface temperature for climate from Along-Track Scanning Radiometers. J Geophys Res-Oceans, 2012, 117(C12): C12013

[28] MerchantCJ, HarrisAR, MaturiE, MacCallumS. Probabilistic physically based cloud screening of satellite infrared imagery for operational sea surface temperature retrieval. Q J R Meteorol Soc, 2005, 131(611): 2735-2755

[29] MurtaghF, BarretoD, MarcelloJ. Decision boundaries using Bayes factors: the case of cloud masks. IEEE Trans Geosci Remote Sens, 2003, 41: 2952-2958

[30] PoulsenC, EgedeU, RobbinsD, SandefordB, TaziK, ZhuT. Evaluation and comparison of a machine learning cloud identification algorithm for the SLSTR in polar regions. Remote Sens Environ, 2020, 248: 111999

[31] Ronneberger O, Fischer P, Brox T (2015) U-Net: convolutional networks for biomedical image segmentation. In: MICCAI 2015, Munich, pp 234–241. https://doi.org/10.1007/978-3-319-24574-4_28

[32] RossowWB, GarderLC. Cloud detection using satellite measurements of infrared and visible radiances for ISCCP. J Climate, 1993, 6(12): 2341-2369

[33] SaundersRW, KriebelKT. An improved method for detecting clear sky and cloudy radiances from AVHRR data. Int J Remote Sens, 1988, 9(1): 123-150

[34] SedanoF, KempeneersP, StroblP, KuceraJ, VogtP, SeebachL, et al. . A cloud mask methodology for high resolution remote sensing data combining information from high and medium resolution optical sensors. ISPRS-J Photogramm Remote Sens, 2011, 66(5): 588-596

[35] Segal-RozenhaimerM, LiA, DasK, ChirayathV. Cloud detection algorithm for multi-modal satellite imagery using convolutional neural-networks (CNN). Remote Sens Environ, 2020, 237

[36] SrivastavaA, JhaD, ChandaS, PalU, JohansenH, JohansenD, et al. . MSRF-Net: a multi-scale residual fusion network for biomedical image segmentation. IEEE J Biomed Health Inform, 2022, 26(5): 2252-2263

[37] SunL, MiX, WeiJ, WangJ, TianXP, YuHY, et al. . A cloud detection algorithm-generating method for remote sensing data at visible to short-wave infrared wavelengths. ISPRS-J Photogramm Remote Sens, 2017, 124: 70-88

[38] UddstromMJ, GrayWR, MurphyR, OienNA, MurrayT. A Bayesian cloud mask for sea surface temperature retrieval. J Atmos Ocean Technol, 1999, 16(1): 117-132

[39] WangL, WangQ. Fast spatial-spectral random forests for thick cloud removal of hyperspectral images. Int J Appl Earth Obs Geoinf, 2022, 112

[40] WeiJ, HuangW, LiZ, SunL, ZhuXL, YuanQQ, et al. . Cloud detection for Landsat imagery by combining the random forest and superpixels extracted via energy-driven sampling segmentation approaches. Remote Sens Environ, 2020, 248

[41] WuK, XuZ, LyuX, RenP. Cloud detection with boundary nets. ISPRS-J Photogramm Remote Sens, 2022, 186: 218-231

[42] YangY, LiP. Scene- and pixel-level analysis of Landsat cloud coverage and image acquisition probability in South and Southeast Asia. Int J Appl Earth Obs Geoinf, 2023, 123

[43] YaoX, GuoQ, LiA. Light-weight cloud detection network for optical remote sensing images with attention-based DeeplabV3+ architecture. Remote Sens, 2021, 13(18)

[44] ZhangGB, GaoXJ, YangJH, YangYW, TanML, XuJ, et al. . A multi-task driven and reconfigurable network for cloud detection in cloud-snow coexistence regions from very-high-resolution remote sensing images. Int J Appl Earth Obs Geoinf, 2022, 114

[45] ZhangWD, JinSL, ZhouL, XieXW, WangFY, JiangLL, et al. . Multi-feature embedded learning SVM for cloud detection in remote sensing images. Comput Electr Eng, 2022, 102