多源多时相遥感影像相对辐射归一化方法研究
作者简介:黄启厅(1983-),男,广西南宁人,博士生,研究方向为遥感信息提取。E-mail:huangqiting830112@163.com
收稿日期: 2015-12-16
要求修回日期: 2016-04-06
网络出版日期: 2016-05-10
基金资助
国家高技术研究发展计划项目(2015AA123901)
广西科学研究与技术开发计划项目(14125008-1-6)
国家高分辨率对地观测系统重大专项(03-Y30B06-9001-13/15-01)
A Study on the Relative Radiometric Normalization of Multi-sources and Multi-temporal Remote Sensing Data
Received date: 2015-12-16
Request revised date: 2016-04-06
Online published: 2016-05-10
Copyright
为满足多源、多时相遥感影像定量化信息提取的应用需求,本文发展了一种半自动化的相对辐射归一化方法。将多源遥感影像的相对辐射归一化分为传感器辐射校正与针对光照等外部因素的辐射归一化2个过程。首先,基于晴空影像,采用分类回归的方式获取传感器辐射校正系数;然后,利用样本传递再分类的方法实现多源影像的半自动分类和传感器辐射偏差校正;最后,基于NDVI差值直方图和类别约束的PIFs自动选取方法,实现影像的相对辐射归一化。采用准同期的GF1-WVF1和Landsat8-OLI影像以及多源时序影像对方法进行了验证,结果表明,本文方法可以对传感器间的辐射偏差进行有效纠正,并在整体上获得比传统方法更好的辐射归一化精度;同时,多源时序影像的辐射校正结果也表明,本文方法能够有效地消除时序影像间的辐射特征波动,使植被等地类的季相变化信息得到更准确地表达,为多源时序影像的协同利用提供了借鉴方法。
黄启厅 , 覃泽林 , 曾志康 . 多源多时相遥感影像相对辐射归一化方法研究[J]. 地球信息科学学报, 2016 , 18(5) : 606 -614 . DOI: 10.3724/SP.J.1047.2016.00606
In order to meet the demand for quantitative information extraction from multi-sources and multi-temporal satellite data, a semi-automatically relative radiometric normalization approach was developed in this paper. The relative radiometric normalization procedure of multi-sources image is divided into two parts: the first one is sensors’ radiometric correction and the other is the normalization of radiometric discrepancy caused by external factors such as the changes of illumination. Firstly, the radiometric correction coefficients of sensors were obtained by the classes-specified regression method based on clear-sky images. Secondly, by applying the sample transferring method, the multi-sources images were semi-automatically classified, and the radiometric deviations of the corresponding sensors were adjusted. Lastly, based on the images’ classification results, the PIFs were automatically chosen by combining the NDVI difference histogram with the class restraint, and the relative radiometric normalization of images was thereby accomplished. A pair of quasi-synchronous GF1-WVF1 and Landsat 8 images, and a set of multi-sources and time-series images were adopted to demonstrate the validity of the presented approach. The results showed that, this method can effectively correct sensors’ radiometric discrepancy and achieved a higher radiometric normalization accuracy when compared with the traditional method.Meanwhile, the results from time-series dataset also demonstrated that this method can effectively reduce the fluctuation of radiation features between time-series images, and the seasonal changes of vegetation were therefore accurately presented. It can thereby offer a reference to the synergic utilization of multi-sources and time-series images.
Fig. 1 Flowchart of relative radiance correction for multi-sources and multi-temporal images图1 多源多时相影像相对辐射归一化流程图 |
Fig. 2 Clear-sky images of the reference and target sensors图2 基准传感器与待纠正传感器重叠区晴空影像(RGB组合) |
Tab. 1 Regression models of GF1-WFV1 sensor with respect to Landsat8-OLI表1 以Landsat8-LIO为基准的GF-WFV1回归模型 |
传感器 | 波段 | 回归方程 | |||
---|---|---|---|---|---|
植被 | 裸地 | 居民地 | 水体 | ||
GF1-WFV1 | B1 | Y =0.8620 X -1.5301 | Y =0.8141 X +4.7266 | Y =0.7349 X +11.473 | Y =0.8109 X +10.305 |
B2 | Y =0.8440 X -3.7146 | Y =0.8368 X +1.7295 | Y =0.8038 X +0.4878 | Y =0.9047 X +1.1441 | |
B3 | Y =0.9149 X -1.4747 | Y =0.8401 X +6.4448 | Y =0.9203 X +0.4012 | Y =0.8078 X +9.5268 | |
B4 | Y =0.7066 X +4.3729 | Y =0.6486 X +7.1556 | Y =0.8062 X +3.9328 | Y =0.6365 X +4.4763 |
Fig. 3 The NDVI discrepancies and the correction results of different sensors图3 不同传感器的NDVI差异及校正结果 |
Tab. 2 Mean radiance of each band before and after the radiometric normalization processed by the proposed method表2 本文方法辐射归一化前后影像各波段的平均辐亮度 |
波段 | 植被 | 裸地 | 居民地 | 水体 | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
GF1-WFV1 | LT8-OLI | 校正影像 | GF1-WFV1 | LT8-OLI | 校正影像 | GF1-WFV1 | LT8-OLI | 校正影像 | GF1-WFV1 | LT8-OLI | 校正影像 | |
B1 | 66.51 | 54.82 | 56.01 | 84.32 | 73.37 | 75.67 | 97.53 | 83.15 | 86.75 | 79.05 | 74.41 | 76.91 |
B2 | 57.33 | 43.68 | 45.47 | 76.22 | 65.35 | 66.95 | 82.43 | 66.74 | 70.63 | 65.06 | 60.01 | 62.14 |
B3 | 29.46 | 25.48 | 26.67 | 74.38 | 68.93 | 72.87 | 58.25 | 54.01 | 56.01 | 34.31 | 37.24 | 36.03 |
B4 | 88.47 | 66.87 | 70.25 | 85.32 | 62.49 | 65.64 | 67.02 | 50.11 | 57.97 | 25.36 | 20.62 | 22.97 |
Mean-Radiance | 60.44 | 47.71 | 49.6 | 80.06 | 67.54 | 70.28 | 76.31 | 63.50 | 67.84 | 50.95 | 48.07 | 49.51 |
注:Mean-Radiance为4个波段的平均辐亮度,单位为W/m2·μm·sr |
Fig. 4 Raw images and radiometric normalization results with respect to different methods图4 原始影像和不同方法的辐射归一化结果(RGB组合) |
Tab. 3 The RMSE of the radiometric normalization results with respect to different methods表3 不同方法辐射归一化结果的RMSE |
波段 | 植被 | 裸地 | 居民地 | 水体 | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
未校正前 | IR | 本文方法 | 未校正前 | IR | 本文方法 | 未校正前 | IR | 本文方法 | 未校正前 | IR | 本文方法 | |
B1 | 10.79 | 3.52 | 1.44 | 11.40 | 13.35 | 2.86 | 15.26 | 17.64 | 4.44 | 5.51 | 4.92 | 2.88 |
B2 | 12.87 | 5.22 | 2.30 | 11.34 | 12.94 | 2.83 | 16.37 | 17.99 | 4.25 | 5.95 | 4.27 | 3.09 |
B3 | 4.25 | 2.64 | 1.48 | 7.22 | 9.27 | 4.21 | 6.51 | 9.33 | 4.87 | 4.44 | 4.12 | 3.24 |
B4 | 16.44 | 8.89 | 3.59 | 23.51 | 26.33 | 3.64 | 17.56 | 19.41 | 8.94 | 6.48 | 5.28 | 3.34 |
Mean-RMSE | 11.09 | 5.07 | 2.20 | 13.37 | 15.47 | 3.39 | 13.93 | 16.09 | 5.62 | 5.6 | 4.65 | 3.15 |
注:未校正前表示归一化前2个传感器辐亮度影像的RMSE;Mean-RMSE为4个波段辐亮度RMSE的平均值,单位为W/m2·μm·sr |
Fig. 5 NDVI curves before radiometric normalization图5 辐射归一化前NDVI曲线 |
Fig. 6 NDVI curves after radiometric normalization图6 辐射归一化后NDVI曲线 |
The authors have declared that no competing interests exist.
[1] |
[
|
[2] |
[
|
[3] |
|
[4] |
|
[5] |
|
[6] |
|
[7] |
|
[8] |
[
|
[9] |
|
[10] |
|
[11] |
|
[12] |
[
|
[13] |
[
|
[14] |
|
[15] |
[
|
[16] |
[
|
[17] |
[
|
[18] |
|
/
〈 | 〉 |