论文部分内容阅读
本文提出了混合像元的概念，研究了基于混合像元的遥感图像分类问题，根据最小二乘法的原理导出了混合像元的分类算法。实验表明：在多光谱图像分类中考虑混合像元的客观存在，可以大大提高遥感图像的分类精度。
In this paper, the concept of mixed pixels is proposed. The remote sensing image classification problem based on mixed pixels is studied. According to the principle of least squares, the classification algorithm of mixed pixels is derived. Experiments show that the objective existence of mixed pixels in multi-spectral image classification can greatly improve the classification accuracy of remote sensing images.