【摘要】 北京持续推进增彩延绿科技示范工程，彩叶植物在城市园林建设和人居环境改善方面发挥着越来越重要的作用。如果能够利用高光谱技术实现快速、无损地观测城市彩叶植物区域分布特点及其生长特征变化，可为进一步优化城市彩叶植物布局，加快城市彩叶植物系统建设提供重要理论依据和数据支撑。高光谱遥感技术的快速发展，不仅提供了大量地被植物光谱信息，而且也提高了光谱分辨率及其响应范围。植物光谱具有一系列特征吸收谱带，能够指示不同树种间的差异，是高光谱进行树种识别的基础。以北京城市常见不同色系彩叶植物15种为研究对象，运用SR-3501便携式地物光谱仪分析了不同色系植物叶片秋季高光谱反射曲线特征；通过对原始光谱数据进行微分变换和特征参数提取，进一步研究了不同色系植物反射特征波段及特征参数差异性和变化规律。结果表明：大叶黄杨具备典型绿色植被光谱曲线特征，即呈“峰”和“谷”的变化特征，紫色系植物表现为同绿色系植物近似的光谱反射特征，红色系植物与黄色系植物光谱反射特征相似；从光谱吸收特征参数角度分析，不同色系植物绿峰/红峰位置表现为红色系植物>紫色系植物>黄色系植物>绿色系植物，而绿峰/红峰反射率、红谷位置和红谷反射率均表现为黄色系植物>红色系植物>紫色系植物>绿色系植物；不同色系植物叶片光谱三边特征参数具有一定的规律性，三边参数可以作为区分不同彩色叶植物与绿色系植物的特征参数，其中红边幅值与红边面积、黄边幅值与黄边面积、蓝边幅值与蓝边面积可分别作为紫色系植物、红色系植物与黄色系植物区别于其他色系植物的重要光谱特征参数。更多还原

【Abstract】 Along with promoting the color extension green technology demonstration project in Beijing, color-leaf plants play an increasingly prominent role in urban landscape construction and improvement of the living environment, especially in recent years. If the regional distribution and growth characteristics of urban colorful leaf plants can be observed quickly and lossless by using hyperspectral technology, important theoretical basis and data support can be provided for further optimizing the layout of urban colorful leaf plants and accelerating the construction of urban color-leaf plants system. In recent years, the rapid development of hyperspectral remote sensing technology provides a lot of ground cover plant spectral information and improves the spectral resolution and response range. Plant spectrum has a series of characteristic absorption bands, which can indicate the differences between different tree species, and is the basis of hyperspectral tree species identification. This paper selected 15 species of colorful leaf plants with different color systems in Beijing as the research object. Moreover, the SR-3501 portable surface feature spectrometer was used to analyze the characteristics of the hyperspectral reflection curve of leaves of plants of different color families in autumn. The difference and variation of the characteristic bands and characteristic parameters of plants of different color families were further studied through the differential transformation and feature parameter extraction. The results showed that Euonymus japonzcus had the characteristics of typical green vegetation spectral curve, which were the changes of “peak” and “valley”; the spectral reflection characteristic of purple leaf plants was similar to that of green plants; the spectral reflection characteristic of red leaf plants was similar to that of yellow leaf plants. Based on spectral absorption characteristic parameters, the green/red peak position of different color plants showed a trend of red leaf plants>purple leaf plants>yellow leaf plants>green leaf plants, and green/red peak reflectivity, red valley location and red valley reflectivity were all represented by the yellow leaf plants>red leaf plants>purple leaf plants>green leaf plants. The characteristic spectrum parameters of the three sides of different color plants had certain regularity and can be used as the characteristic parameters to distinguish the different colored plants with green plants. In comparison, the red amplitude and red edge area, yellow amplitude and yellow edge area, blue amplitude and blue edge area could be used as the important spectral characteristic parameters to distinguish the purple leaf plants, the red leaf plants and the yellow leaf plants from green plants. This study provides a theoretical basis for applying the hyperspectral technique in the future observation of urban color leaf plant system construction.更多还原