论文部分内容阅读
GEOBIA(Geographic Object-Based Image Analysis)技术针对高空间分辨率遥感影像分析的效果和精度远优于基于像元的传统方法。影像分割作为GEOBIA中的关键技术,学者们对此已经做了大量的研究,提出众多分割算法。对分割算法进行评价和分割技术本身同样重要,通过分割评价可以对分割算法的性能进行评价,比较不同分割算法的优劣,为影像选择合适的分割算法并设定合适的分割参数。影像分割的目的是为了实现影像分析操作的自动化,而主观评价法、系统评价法和分析评价法,因其无法给出客观定量指标的特点,难以应用于实时、自动化的高分辨率影像信息提取与分析系统当中。加之近年来针对分割评价方法的研究远远落后于分割算法本身,因此对定量分割评价方法进行综述对于影像分割方法及其应用研究意义重大。本文对现有的评价方法进行系统总结,建立了针对高空间分辨率遥感影像分割评价方法的分类体系。对各种方法,特别是定量的实验评价法进行对比,分析其应用范围和优劣,最后指出了高空间分辨率遥感影像分割评价未来的改进方向和应用前景。
GEOBIA (Geographic Object-Based Image Analysis) technology is superior to the traditional pixel-based method for high spatial resolution remote sensing image analysis. Image segmentation as the key technology in GEOBIA, scholars have done a lot of research on this, put forward many segmentation algorithms. It is equally important to evaluate and segment the segmentation algorithm itself. The performance of the segmentation algorithm can be evaluated through the segmentation and evaluation. The advantages and disadvantages of different segmentation algorithms are compared and the suitable segmentation algorithm is selected and the appropriate segmentation parameters are set. The purpose of image segmentation is to automate the image analysis operations. Subjective evaluation, systematic evaluation and analytical evaluation can not be applied to real-time and automatic high-resolution image information extraction because of their inability to give objective quantitative indicators And analysis system. In addition, the researches on segmentation evaluation methods lag far behind the segmentation algorithms in recent years. Therefore, reviewing the quantitative segmentation assessment methods is of great significance to the image segmentation methods and their applications. In this paper, the existing evaluation methods are systematically summarized, and a classification system for the high spatial resolution remote sensing image segmentation evaluation method is established. The comparison of various methods, especially the quantitative experimental evaluation method, analyzes its application range and advantages and disadvantages. Finally, it points out the future improvement direction and application prospect of high spatial resolution remote sensing image segmentation.