| 张翼飞,朱建华,田震,贾迪,高飞.近岸水体异源遥感反射率产品的融合方法研究[J].海洋通报,2022,(4): |
| 近岸水体异源遥感反射率产品的融合方法研究 |
| Study on fusion method of remote sensing reflectance products from different sources in nearshore water |
| 投稿时间：2022-03-30 修订日期：2022-05-11 |
| DOI：10.11840/j.issn.1001-6392.2022.04.001 |
| 中文关键词: 数据融合 生物光学模型 混合像元分解 小波变换 哨兵卫星 遥感反射率 |
| 英文关键词:fusion bio-optical model unmixing wavelet transform Sentinel satellite reflectance |
| 基金项目:新一代水色卫星多载荷数据融合技术研究（K4210C020） |
|
| 摘要点击次数: 151 |
| 全文下载次数: 26 |
| 中文摘要: |
| 数据融合能够综合利用多源遥感数据优势，获取高空间分辨率和高测量精度的遥感影像，这对于近岸水体生态环境监测和灾害预警均具有重要意义。但目前使用的数据融合方法多是针对内陆水体或大洋水体的，其在近岸水体的适用性仍需进一步评估，因而本文开展了近岸水体遥感数据融合方法对比研究。本文以高空间分辨率的Sentinel-2 MSI影像和中等空间分辨率的Sentinel-3 OLCI影像为数据源，分别开展了基于小波变换（WTBF）、生物光学模型（BOBF）和混合像元分解（IUBF）三种方法的数据融合实验，并在此基础上借助现场实测Rrs数据与融合影像对应点的平均相对误差MRE、均方根误差RMSE、偏差bias和平均梯度对各种方法的融合数据质量和区域适用性进行了评估。结果表明：（1）OLCI的Rrs数据精度高于MSI。其中MSI在443 nm、560 nm和665 nm三个波段的MRE、RMSE均高于OLCI，表明MSI的Rrs精度相对OLCI较低；目视效果和平均梯度表明MSI的清晰度高于OLCI；（2）BOBF是三种方法中融合效果最佳的算法。WTBF和BOBF生成融合影像在三个波段的MRE、RMSE优于MSI。综合MRE、RMSE和bias来看，BOBF和WTBF所生成的影像Rrs准确性高于MSI，而IUBF生成的融合影像的MRE、RMSE和bias相对WTBF和BOBF较高，准确性较差；目视评价和平均梯度表征WTBF和BOBF生成影像的清晰度与MSI相近，BOBF清晰度高于WTBF，IUBF生成影像清晰度相对于OLCI有所提高但未达到MSI水平；（3）在Rrs更低的烟台近岸海域，BOBF生成的融合影像清晰度与MSI相当且Rrs准确性相对于MSI更高，BOBF在该海域具有良好的适用性。 |
| 英文摘要: |
| Data fusion can comprehensively utilize the advantages of multi-source remote sensing data to obtain images with high spatial resolution and high measurement accuracy, which is of great significance for coastal water ecological environment monitoring and disaster warning. However, due to the complex composition of nearshore water bodies, the applicability of existing data fusion methods still needs to be further evaluated. Therefore, a comparative study of nearshore water remote sensing data fusion methods is carried out in this paper. In this paper, Sentinel-2 MSI with high spatial resolution and Sentinel-3 OLCI with medium spatial resolution are used as data sources. Based on this data source, three data fusion experiments (wavelet transform based fusion(WTBF), bio-optical based fusion(BOBF), and improved unmixing based fusion(IUBF)) are carried out. On this basis, the quality of the fusion data is evaluated with the in-situ Rrs data, and the regional applicability is evaluated. The results show that :(1) the Rrs data accuracy of OLCI is higher than that of MSI. The mean relative error (MRE), root mean square error (RMSE) of MSI at 443 nm, 560 nm and 665 nm are higher than those of OLCI, indicating that The Rrs accuracy of MSI is low. (2) The Rrs accuracy of BOBF and WTBF images is higher than that of MSI，the clarity is higher than that of OLCI, and BOBF is the algorithm with the best fusion effect among the three methods. The MRE, RMSE and bias of fusion images generated by WTBF and BOBF are better than those of MSI, while the MRE, RMSE and bias of fusion images generated by IUBF are lower than those of WTBF and BOBF, and the accuracy is poor. The clarity of images generated by WTBF and BOBF was similar to that of MSI by visual evaluation and mean gradient. The clarity of images generated by IUBF was improved compared with OLCI but did not reach the level of MSI. (3) In the nearshore area of Yantai with lower Rrs, the resolution of fusion image generated by BOBF is equal to that of MSI and the accuracy of Rrs is higher than that of MSI, indicating that BOBF has good applicability in this area. |
| 查看全文 下载PDF阅读器 |
| 关闭 |
|
|
|