城市绿地衰老退化特征提取仿真

Urban Green Space Aging Degradation Characteristics Extraction Simulation

对城市绿地衰老退化特征的提取,能够有效改善城市环境,净化居住区空气。对绿地衰老退化特征提取问题的研究,需要将绿地图像分割为多个特定区域,给出退化特征聚类的目标函数,完成对城市绿地衰老特征的提取。传统方法先对绿地图像进行预处理,将直方图的绿地衰退本身部分变换成百分率直方图,但忽略了给出绿地退化特征聚类的目标函数,导致提取精度偏低。提出基于颜色矩的城市绿地衰老退化特征提取方法。上述方法先利用颜色完全局部二值模式对采集的城市绿地图像进行预处理,提取绿地图像纹理特征,利用HSV颜色直方图提取图像颜色特征,将颜色和纹理特征相融合,采用模糊C-均值聚类方法将绿地图像分割为多个特定的、具有独特性质的区域,给出城市绿地衰老退化特征聚类的目标函数,提取城市绿地衰老退化特征。仿真证明,所提方法特征提取精度较高。

This research proposes a feature extraction method for senescent degeneration of urban green space based on color moment. First of all, the research pretreated image of urban green space collected by utilizing local bi- nary patterns of full color and extracted texture feature of green image, and then extracted image color feature by using HSV color histogram to fuse color and texture feature. Moreover, the research divided green image into several specif- ic and distinctive quality regions using fuzzy C-mean value clustering method. At last, the research provided objec- tive function of feature clustering of senescent degeneration of urban green space and extracted senescent degeneration feature for urban green space. The simulation results demonstrate that the method in this paper has higher extraction accuracy.