亚像元制图适应性分析与评价——以天津市津南区和北京市海淀区土地覆被制图为例
作者简介:江昱(1990-),女,安徽合肥人,硕士生,主要从事遥感亚像元制图应用研究。E-mail: jiangy@lreis.ac.cn
收稿日期: 2014-12-08
要求修回日期: 2015-02-27
网络出版日期: 2015-10-10
基金资助
国家自然科学基金项目(41471296);国家科技支撑计划课题(2012BAH33B01)
Reliability Analysis and Assessment of Sub-Pixel Mapping: A Case Study with Landsat-5Image and HJ-1A Image Based on VBSPM
Received date: 2014-12-08
Request revised date: 2015-02-27
Online published: 2015-10-10
Copyright
亚像元制图作为一种降尺度分类方法,可利用低分辨率影像获取高分辨率分类图。本文旨在探讨亚像元制图的降尺度分类结果与高分辨率影像分类精度和分类特征上的一致性。实验以天津市津南区和北京市海淀区为研究区,分别对中空间分辨率影像(TM或HJ)进行亚像元制图和对高空间分辨率影像(ALOS或ZY)进行硬分类得到相同空间分辨率的分类结果,从绝对精度、相对精度、空间结构和空间格局上,对2幅分类结果进行分析和评价。实验结果显示:(1)分类精度上,TM和HJ影像的亚像元制图结果,以地面验证样本为参考的绝对总体精度分别为84%和82%,以高分辨率影像(ALOS和ZY影像)硬分类结果,为参考的相对总体精度分别为82%和77%;(2)分类特征上,中空间分辨率影像亚像元制图结果的空间相关性较强、斑块数量较少、聚集度较高,但与高分辨率影像分类结果的总体结构相似,各类别的面积比例基本一致。因此,亚像元制图结果在分类精度和分类特征上与高空间分辨率影像分类结果具有较强的一致性,在缺少高分辨率土地覆被制图时,可将亚像元制图获取的降尺度分类图作为替代数据。
江昱 , 葛咏 , 陈跃红 , 宋海荣 , 胡建龙 . 亚像元制图适应性分析与评价——以天津市津南区和北京市海淀区土地覆被制图为例[J]. 地球信息科学学报, 2015 , 17(10) : 1215 -1223 . DOI: 10.3724/SP.J.1047.2015.01215
Some high-resolution land cover maps are not free or available for direct use due to its economic value, the impact of weather or its confidentiality. As a downscaling classification method, sub-pixel mapping (SPM) can produce classification data with spatial resolutions finer than the original input data. We aim to explore the consistency between SPM results and classification data extracted from high-resolution remote sensing images on their accuracy and spatial characteristics. Two experiments were performed: one is in Jinnan District, Tianjin City with Landsat-5 TM image, and the other is in Haidian District, Beijing City with HJ image. Results show that the overall absolute accuracies of SPM results produced by TM and HJ images are 84% and 82% respectively. The overall relative accuracies of Landsat-5 and HJ SPM results were 82% and 77% by taking high-resolution classifications as reference. Furthermore, the overall structures and proportions based on the results using the proposed method are similar with high-resolution classifications. Therefore, with the absence of high-resolution land cover map, results generated by SPM could provide an alternative for land cover data source.
Key words: remote sensing classification; sub-pixel mapping; adaptability; ALOS; Landsat TM
Tab. 2 Main parameters of the experiment data表2 实验数据的主要参数 |
影像 | 获取时间 | 空间分辨率(m) | 波段(mm) | 影像大小 |
---|---|---|---|---|
TM | 2009-10-17 | 30 | B1: 0.45~0.52 B2: 0.52~0.60 B3: 0.63~0.69 B4: 0.76~0.90 | 818×860 |
ALOS | 2009-10-17 | 10 | B1: 0.42~0.50 B2: 0.52~0.60 B3: 0.61~0.69 B4: 0.76~0.89 | 2454×2580 |
HJ | 2012-04-16 | 30 | B1: 0.43~0.52 B2: 0.52~0.60 B3: 0.63~0.69 B4: 0.76~0.90 | 999×937 |
ZY | 2012-04-17 | 6 | B1: 0.45~0.52 B2: 0.52~0.59 B3: 0.63~0.69 B4: 0.77~0.89 | 4995×4685 |
Fig. 1 Flow chart of the experiment图1 实验流程图 |
Fig. 3 Locations of the study area图3 实验区位置示意图 |
Fig. 4 Standard false color composite of the experiment images图4 实验影像的标准假彩色合成图像 |
Tab. 3 Jeffries-Matusita distance among the training samples of each class for from experiment images表3 实验影像中各类别训练样本之间的J-M距离 |
水体/建筑 | 水体/植被 | 水体/裸地 | 建筑/植被 | 建筑/裸地 | 植被/裸地 | ||
---|---|---|---|---|---|---|---|
实验1 | TM | 1.922 | 1.991 | 2.000 | 1.945 | 1.993 | 1.861 |
ALOS | 1.927 | 1.995 | 2.000 | 1.952 | 1.976 | 1.762 | |
实验2 | HJ | 1.995 | 1.986 | 2.000 | 1.993 | 1.952 | 2.000 |
ZY | 1.963 | 2.000 | 2.000 | 1.998 | 1.901 | 2.000 |
Fig. 5 Soft classification results of the experiments图5 实验软分类结果 |
Fig. 6 Sub-pixel mapping results and hard classification maps图6 亚像元制图结果和硬分类结果 |
Tab. 4 Absolute accuracy assessment of sub-pixel mapping results and classification images表4 亚像元制图结果和硬分类结果的绝对精度评价结果 |
实验区 | 影像 | OA(%) | KP | 水体 | 建筑 | 植被 | 裸地 | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
PA(%) | UA(%) | PA(%) | UA(%) | PA(%) | UA(%) | PA(%) | UA(%) | |||||||
1 | ALOS | 85.67 | 0.774 | 80.00 | 87.61 | 87.62 | 83.61 | 91.85 | 92.19 | 42.86 | 45.00 | |||
TM | 84.27 | 0.752 | 76.84 | 82.49 | 78.50 | 87.96 | 94.44 | 90.37 | 51.08 | 39.86 | ||||
2 | ZY | 86.60 | 0.779 | 96.23 | 80.95 | 86.81 | 91.11 | 88.86 | 93.93 | 78.52 | 59.34 | |||
HJ | 82.11 | 0.691 | 39.62 | 70.00 | 87.08 | 83.75 | 87.47 | 82.04 | 51.73 | 73.20 |
注:OA为总体精度;KP为Kappa系数;PA为生产者精度;UA为用户精度 |
Tab. 5 Relative accuracy assessment of sub-pixel mapping and hard classification results derived from low spatial resolution remote sensing images表5 中分辨率影像的亚像元制图结果和硬分类结果的相对精度评价结果 |
影像 | OA(%) | KP | 水体 | 建筑 | 植被 | 裸地 | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
PA(%) | UA(%) | PA(%) | UA(%) | PA(%) | UA(%) | PA(%) | UA(%) | ||||||
TM_SPM | 82.40 | 0.724 | 79.26 | 81.82 | 78.80 | 89.16 | 91.92 | 89.27 | 33.68 | 23.39 | |||
TM_hard | 82.09 | 0.718 | 78.87 | 81.05 | 78.56 | 88.71 | 91.73 | 88.80 | 32.16 | 23.01 | |||
HJ_SPM | 76.67 | 0.604 | 31.75 | 88.04 | 82.69 | 77.25 | 85.60 | 77.14 | 42.57 | 70.36 | |||
HJ_hard | 76.61 | 0.604 | 31.81 | 80.81 | 82.42 | 77.36 | 85.64 | 76.95 | 43.09 | 70.47 |
注:OA为总体精度;KP为Kappa系数;PA为生产者精度;UA为用户精度 |
Tab. 7 Statistics of spatial structure characteristics for each class in the classification results表7 分类结果中各类别地物的纹理特征统计 |
影像 | 类别 | 均一性 | 对比度 | 相异性 | 信息熵 | 二阶矩 | 相关性 | |
---|---|---|---|---|---|---|---|---|
实验1 TM | 水体 | 0.98 | 71.48 | 1.13 | 0.05 | 0.97 | 0.97 | |
建筑 | 0.96 | 143.71 | 2.28 | 0.1 | 0.94 | 0.93 | ||
植被 | 0.96 | 143.71 | 2.28 | 0.1 | 0.94 | 0.93 | ||
裸地 | 0.96 | 150.15 | 2.38 | 0.1 | 0.94 | 0.92 | ||
ALOS | 水体 | 0.97 | 108.59 | 1.72 | 0.07 | 0.95 | 0.94 | |
建筑 | 0.94 | 227.84 | 3.62 | 0.15 | 0.91 | 0.87 | ||
植被 | 0.94 | 242.39 | 3.85 | 0.16 | 0.9 | 0.86 | ||
裸地 | 0.95 | 187.61 | 2.98 | 0.13 | 0.93 | 0.86 | ||
实验2 HJ | 水体 | 0.99 | 11.28 | 0.18 | 0.00 | 0.99 | 0.00 | |
建筑 | 0.98 | 90.70 | 1.43 | 0.06 | 0.96 | 0.00 | ||
植被 | 0.98 | 68.44 | 1.09 | 0.04 | 0.97 | 0.00 | ||
裸地 | 0.97 | 106.49 | 1.69 | 0.07 | 0.96 | 0.00 | ||
ALOS | 水体 | 0.99 | 5.61 | 0.09 | 0.00 | 0.99 | 0.00 | |
建筑 | 0.92 | 310.77 | 4.93 | 0.20 | 0.87 | 0.00 | ||
植被 | 0.94 | 215.96 | 3.43 | 0.14 | 0.91 | 0.00 | ||
裸地 | 0.97 | 113.92 | 1.81 | 0.08 | 0.95 | 0.00 |
4.2.2 格局特征评价 |
Tab. 8 Statistics of landscape pattern indexes for the classification results and for each class表8 分类结果及各类别地物的景观格局指数统计 |
水平 | 指数 | PD | LSI | COHESION | PLADJ | SHDI | SHEI | ||
---|---|---|---|---|---|---|---|---|---|
实验1 | 景观 | TM | 39 | 58 | 100 | 80 | 1.17 | 0.84 | |
ALOS | 142 | 101 | 100 | 83 | 1.14 | 0.82 | |||
类别 | TM | 水体 | 19 | 62 | 98 | 90 | - | - | |
建筑 | 17 | 63 | 100 | 95 | - | - | |||
植被 | 8 | 171 | 91 | 70 | - | - | |||
裸地 | 6 | 69 | 100 | 93 | - | - | |||
ALOS | 水体 | 16 | 109 | 95 | 83 | - | - | ||
建筑 | 90 | 119 | 99 | 91 | - | - | |||
植被 | 5 | 292 | 74 | 38 | - | - | |||
裸地 | 22 | 135 | 100 | 88 | - | - | |||
实验2 | 景观 | HJ | 76 | 80 | 100 | 95 | 0.95 | 0.69 | |
ALOS | 97 | 104 | 100 | 94 | 1.06 | 0.77 | |||
类别 | HJ | 水体 | 7 | 30 | 96 | 90 | - | - | |
建筑 | 22 | 103 | 100 | 96 | - | - | |||
植被 | 31 | 94 | 100 | 95 | - | - | |||
裸地 | 16 | 81 | 98 | 92 | - | - | |||
ZY | 水体 | 6 | 49 | 95 | 89 | - | - | ||
建筑 | 18 | 128 | 100 | 95 | - | - | |||
植被 | 29 | 85 | 100 | 96 | - | - | |||
裸地 | 45 | 154 | 97 | 88 | - | - |
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] |
|
[19] |
|
[20] |
|
[21] |
|
[22] |
|
[23] |
|
[24] |
|
[25] |
|
[26] |
|
[27] |
|
[28] |
|
[29] |
|
[30] |
|
[31] |
|
[32] |
|
[33] |
|
[34] |
|
[35] |
|
[36] |
|
/
〈 | 〉 |