基于遥感图像三区光谱特征的水网城市区域规划协调控制方法

Coordinated control method for urban planning of water network based on spectral features of three regions in remote sensing images

摘要:

水网城市区域规划功能区图像相似性特征较强，且受生态环境、水资源调度等多个因素影响，造成水网区域规划协调性较差，提出基于遥感图像三区光谱特征的水网城市区域规划协调控制方法。使用高分辨率遥感设备采集城市影像；运用形态学算法，利用结构树，通过击中、击不中变换的方式对图像做细化处理；从亮度均值、目标标准差和亮度差级指数三方面提取图像光谱特征；采用最邻近分类算法，建立分类模型，以优先功能区、一般功能区和冲突协调区的自身环境和光谱特征为依据，将提取的特征输入到分类模型中，自动输出划分结果，利用 A*算法简化划分步骤，实现水网城市规划协调控制。实验结果表明，所提方法提高了遥感图像质量，分类结果符合不同功能区的特征要求；兼顾其它生态要素，规划具有协调性，规划效果较好；可在2s内完成目标区域规划，规划时间较短；对于目标规划数据的检测误差小、准确率高，大幅度提高搜索效率，能够为水网城市区域规划协调控制提供可靠依据。

Abstract:

The similarity of functional area images of water network urban regional planning is strong, and affected by ecological environment, water resource scheduling and other factors, resulting in poor coordination of water network regional planning. A coordinated control method of water network urban regional planning based on spectral features of three regions of remote sensing image is proposed. Using high resolution remote sensing equipment to collect urban images; Using morphology algorithm and structure tree, the image is refined by means of hit and miss transformation. The spectral features of images were extracted from brightness mean, target standard deviation and brightness difference index. The nearest neighbor classification algorithm was used to establish A classification model. Based on the environment and spectral characteristics of priority functional areas, general functional areas and conflict coordination areas, the extracted features were input into the classification model and the division results were automatically output. A* algorithm was used to simplify the division steps and realize the coordination and control of water network urban planning. The experimental results show that the proposed method improves the quality of remote sensing images, and the classification results meet the feature requirements of different functional areas. Taking into account other ecological elements, the planning is coordinated and the planning effect is good. The target area planning can be completed within 2s, and the planning time is short. The detection error of target planning data is small, the accuracy is high, the search efficiency is greatly improved, and it can provide a reliable basis for the coordination and control of urban planning of water network.