基于蓝噪声理论的遥感图像森林植被分割研究

Remote sensing image segmentation based on the theory of forest vegetation blue noise

研究了森林植被分割经典算法,为有效处理森林植被纹理尺度问题,提出了一种基于蓝噪声理论描述遥感图像森林植被纹理特征的森林植被分割方法。这是一种新的植被纹理刻画方法和纹理尺度计算方法,利用得到的纹理尺度设计特定的结构元进行数学形态学运算,提取森林植被信息,得到分割结果。实验结果验证了新方法的准确性和适应性。

This paper researches divided the forest vegetation classical algorithm, in order to effectively deal with the scale of forest vegetation texture problem, a description of the forest vegetation remote sensing image texture features based on the theory of forest vegetation blue noise segmentation method. This is a new method of vegetation texture characterization and texture scale calculation methods. It uses resulting texture to design specific structural element scale mathematical morphology operation , extract forest cover information, and obtain segmentation results. The experimental results verify the accuracy and adaptability of the new method.