基于震灾图像维度分析下景观带毁损程度评估的探讨
Discussion of Damage Degree Evaluation of Landscape Belts Based on Dimension Analysis of Earthquake Disaster Images

地震灾害是所有自然灾害中破坏程度最为严重的灾害之一，当景观带遭到地震破坏时，景观带中的规则建筑结构发生改变，景观带毁损程度评估是城市避震减灾体系研究的重点内容，因此对景观带中规则建筑毁损程度研究有重要意义。为深入探讨景观带毁损程度评估的问题，基于震灾图像维度分析下对景观带毁损程度的评估方法进行尝试。根据规则建筑震灾图像质量，按图像维度的变化特征对规则建筑结构的破碎程度进行划分，获取破碎概率分布，根据破碎概率划分最佳阈值，采用分数维方法提取图像维度特征，根据提取到的规则建筑震灾图像维度特征，将毁损的规则建筑空间自相关程度转换成正态分布曲线，进而构建评估模型对规则建筑毁损程度进行评估。针对实例分析，实现基于震灾图像维度分析下景观带毁损程度评估方法并进行了探讨。
Earthquakes are one of the most serious of all natural disasters. If a landscape belt is destroyed by an earthquake, regular building structures in the landscape belt will be also changed. Degree-of-damage evaluation of landscape belts is key to urban earthquake avoidance and disaster reduction, so it is of great significance to study degrees-of-damage to regular buildings in the landscape belt. In this paper, a method based in dimensional analysis of earthquake disaster images is used to evaluate the damage degree of landscape belts. According to the earthquake disaster image quality of regular buildings, fragmentation degree was categorized based on the change characteristics of the image dimensions. Optimal threshold was then calculated according to the crushing probability, and the fractional-dimension method was used to extract image dimension features. On this basis, the spatial autocorrelation degree of damaged regular buildings was transformed into a normal distribution curve. A model was then constructed to evaluate the degree-of-damage of the buildings. This case study proved that the method of landscape zone damage degree evaluation, based on dimensional analysis of earthquake disaster images, can be applied to real-life cases