航拍图像道路数据集制备迭代最近边缘算法

Iterative nearest edge algorithm for aerial image road dataset preparation

数据集制备的质量和效率是遥感图像智能处理领域所关注的共性基础问题。针对航拍图像的道路提取数据集制备困难的问题,本文提出了一种迭代最近边缘特征优化算法的航拍图像道路提取数据集制备算法。该算法首先通过人工辅助,建立航拍图像和卫星图像的单应变换关系,将卫星图像投影到航拍图像上,实现基于四点法的卫星图像到航拍图像的粗配准;然后采用边缘检测算子提取粗配准后图像的边缘特征;最后利用基于迭代最近边缘优化算法完成图像的精配准,提高航拍图像的道路数据集制备精度。通过道路提取数据集制备试验,证明了本文所提出的道路数据集制备方法在满足数据集精度要求的同时,能够显著提高道路数据集制备的效率。

The quality and efficiency of dataset preparation are common basic issues that are concerned in the field of remote sensing image intelligent processing. Aiming at the difficulty of preparing the road extraction dataset for aerial image, this paper proposes an iterative algorithm for the optimization of the nearest edge feature to prepare the road extraction dataset for aerial image. The algorithm first establishes the homography transformation relationship between aerial image and satellite image through manual assistance, and projects the satellite image onto the aerial image to realize the coarse registration of the satellite image to the aerial image based on the four-point method. Then, the edge detection operator is used to extract the edge features of the image after the rough registration. Finally, the precise registration of the image is completed by the iterative nearest edge optimization algorithm, which improves the preparation accuracy of the road dataset of the aerial image. At the end of the thesis, a road extraction dataset preparation experiment is carried out, which proves that the road dataset preparation method proposed in this paper can significantly improve the efficiency of road dataset preparation while meeting the accuracy requirements of the dataset.