土地覆盖变化检测中不同相对辐射归一化方法的评价

发布时间：2018-05-02 07:20

本文选题：土地覆盖变化 + 多时相遥感数据　；参考：《南京林业大学学报(自然科学版)》2017年04期

【摘要】：【目的】从土地覆盖变化检测性能的角度寻求一种精度较高且归一化过程相对简单、客观和自动化的辐射归一化方法。【方法】借助2013年7月和2015年9月的南京地区Landsat 8 OLI图像并以2013年影像为基准,执行基于伪不变特征(pseudo-invariant features,PIF)、时不变点群(temporally invariant cluster,TIC)、全景影像回归(wall-to-wall regression,WWR)和加权不变点(weighted invariant pixels,WIP)4种方法,分波段对2015年的多光谱图像进行辐射归一化处理。在此基础上利用变化向量分析技术(change vector analysis,CVA)进行土地覆盖变化信息提取,并借助高空间分辨率Google Maps对提取到的变化信息进行空间一致性验证。【结果】使用PIF、TIC、WWR和WIP方法对图像归一化后,利用变化向量分析提取的土地覆盖变化信息与目视解译得出的土地覆盖变化信息的空间一致性分别为79.63%、81.75%、72.72%和82.59%;依不同变化地类给出变化检测阈值,算法的空间一致性会进一步提高。【结论】考虑到算法的客观性、自动化程度和变化检测精度,在4种方法中,TIC法和WIP法更适合工程化的图像辐射归一化操作。
[Abstract]:[objective] from the point of view of detecting performance of land cover change, it is relatively simple to seek a kind of high precision and normalization process. Objective and automated radiometric normalization methods. [methods] based on the Landsat 8 OLI images of Nanjing in July 2013 and September 2015, Four methods based on pseudo-invariant features, temporal invariant clustering, panoramic image regression wall-to-wall regression and weighted invariant points weighted invariant pixels are implemented. The multispectral images in 2015 are processed by spectral radiometric normalization. On this basis, the change vector analysis technique was used to extract the information of land cover change. The spatial consistency of the extracted change information is verified by Google Maps with high spatial resolution. [results] the image is normalized by using the PIF-TIC-WWR and WIP methods. The spatial consistency of the land cover change information extracted by the change vector analysis and the land cover change information obtained by visual interpretation were 79.63% and 82.59%, respectively. The spatial consistency of the algorithm will be further improved. [conclusion] considering the objectivity of the algorithm, the degree of automation and the accuracy of change detection, the TIC and WIP methods are more suitable for engineering image radiation normalization operation.
【作者单位】：南京林业大学林学院南方现代林业协同创新中心;
【基金】：国家林业局“948”项目(2014-4-25) 国家自然科学基金项目(31670552) 江苏高校优势学科建设工程资助项目(PAPD)
【分类号】：P237

【相似文献】