论文部分内容阅读
光谱维噪声使地物光谱扭曲或变形,中心波长偏移,影响地物信息提取和地表参量反演的精度。对光谱维噪声进行滤波处理,有利于改善遥感数据定量应用的效果。由于数学形态滤波的原理简单且较易实现,被应用到植被光谱以及有机化合物光谱的研究中。运用数学形态滤波对地面实测小麦光谱去噪,一方面对滤波后的光谱进行噪声和波形相似度的直观分析,另一方面通过植被指数反演小麦理化参量进行定量应用评价。结果表明,与传统Savitzky-Golay滤波相比,在可见-近红外波段范围内,数学形态滤波去噪后的光谱能够保持可见—近红外波段原始光谱的固有特征,叶面积指数和叶绿素的反演精度比去噪前有小幅提升,主要原因是实测光谱在该谱段范围的噪声影响很小;在短波红外波段范围内,数学形态滤波能有效去除短波红外大尺度噪声,提高叶片含水量的反演精度。而传统Savitzky-Golay滤波只能削弱短波红外大尺度噪声。广义形态滤波去噪后植被指数和叶片含水量之间的R2最高可达0.5130(去噪前0.3753),叶片含水量的反演值与实测值之间的R2最高可达0.4221(去噪前0.3097),RMSE为0.0243(去噪前0.0318),优于传统Savitzky-Golay滤波。
Spectral dimension noise distorts or distorts the spectrum of the object, shifts the center wavelength, and affects the accuracy of the ground object information extraction and surface parameter inversion. Spectral noise reduction filtering, is conducive to improving the quantitative application of remote sensing data. Because of the simple and easy-to-implement principle of mathematical morphology filtering, it has been applied to the study of vegetation spectrum and organic compound spectrum. Using mathematical morphology filter to measure the spectral noise of wheat measured on the ground, on the one hand, the noise and waveform similarity of the filtered spectrum can be directly analyzed; on the other hand, the quantitative evaluation of wheat physical and chemical parameters can be performed by the vegetation index inversion. The results show that, compared with the traditional Savitzky-Golay filter, in the visible-near-infrared range, the mathematical morphology filter denoising spectrum can maintain the intrinsic characteristics of the original visible-near-IR band, leaf area index and chlorophyll inversion The accuracy is slightly higher than that before denoising, mainly due to the fact that the measured spectrum has little effect on the noise in the spectral range. In the short-wave infrared range, the mathematical morphology filtering can effectively remove large-scale infrared short-wave noise and improve the moisture content of the blade Performing precision. The traditional Savitzky-Golay filter can only weaken the short-wave infrared large-scale noise. The R2 between the vegetation index and the leaf water content after denoising by the generalized morphological filtering can reach 0.5130 (0.3753 before denoising). The R2 between the inversion value of the leaf water content and the measured value can reach 0.4221 (0.3097 before denoising) ), And the RMSE is 0.0243 (0.0318 before denoising), which is better than the traditional Savitzky-Golay filter.