论文部分内容阅读
为了解决仅仅依赖遥感图像光谱特征较难正确区分光谱相近的植被覆盖下垫面林地与农田的问题,对美国Landsat 7 ETM卫星遥感图像进行小波变换分解,研究林地、农田样点的小波系数特征的差异,研究BP人工神经网络识别光谱相近的林地与农田的方法。结果表明,小波变换对遥感图像地物的空间变化非常敏感,在一定尺度下能分离林地与农田地物的细节特点的差异,其多波段小波系数特征存在差异,可以构成光谱相近地物林地、农田的识别特征向量,使用BP神经网络对样本训练后可有效地识别光谱相近的林地与农田地物。
In order to solve the problem that the spectral characteristics of remote sensing images depend on remote sensing images, it is difficult to distinguish the forest land and farmland under the same spectral coverage. The wavelet transform decomposition of the US Landsat 7 ETM satellite remote sensing image is used to study the characteristics of wavelet coefficients of forest land and farmland sample Differences, the study BP artificial neural network to identify similar spectral woodland and farmland approach. The results show that the wavelet transform is very sensitive to the spatial variation of the features of the remote sensing image. The differences of the details of the characteristics of the forest land and the farmland can be separated at a certain scale. The wavelet coefficients of the multi- Farmland identification feature vector, the use of BP neural network training samples can be effectively identified similar spectral woodland and farmland features.