基于形态学运算的建筑垃圾面向对象多特征遥感识别

Identification of Construction Waste Information with Multiple Features using Object-oriented Morphological Operation

随着城市化进展加快,建筑垃圾的科学管理已经成为绿色建筑和新时代环境保护战略下的重要内容。建筑垃圾的长期堆积和不恰当处置会带来一系列环境和社会问题,严重制约城市绿色可持续发展,因此研究快速发现建筑垃圾信息的方法刻不容缓。随着遥感技术迅速发展,高分辨率卫星遥感影像能够实时或准实时反映地表信息,成为提取建筑垃圾信息的新途径。基于建筑垃圾独特的堆积形态、周边环境和影像特征,形态运算可以实现建筑垃圾遥感影像信息增强的目的,进而结合形态、光谱、几何、纹理特征对面向对象的多特征建筑垃圾影像进行层次分类。最后,通过混淆矩阵和可分性的准确性评价,对实验结果进行质量评价和讨论。

With the development of urbanization,the scientific management of construction waste has become an important part of green building and environmental protection strategy in the new era.The long-term accumulation and improper disposal of construction waste will bring about a series of environmental and social problems,which seriously restrict the green and sustainable development of cities.Therefore,it is urgent to study the method of quickly discovering construction waste.With the rapid development of remote sensing technology,high-resolution satellite remote sensing image can reflect the surface information in real time or quasi-real time,which has become a new way to extract construction waste information.Based on the unique accumulation form,surrounding environment and image characteristics of construction waste,morphological operation can realize the purpose of enhancing construction waste information,and then carry out hierarchical classification of object-oriented multi-feature construction waste image by combining morphological,spectral,geometric and texture characteristics.Finally,the quality of the experimental results is evaluated and discussed by the confusion matrix and the accuracy evaluation of separability.