段敏杰, 李新宇, 赵松婷, 等. 不同滞尘环境下植物叶片高光谱特征变化研究[J]. 西南林业大学学报(自然科学), 2020, 40(3): 88–94 . DOI: 10.11929/j.swfu.201908022
引用本文: 段敏杰, 李新宇, 赵松婷, 等. 不同滞尘环境下植物叶片高光谱特征变化研究[J]. 西南林业大学学报(自然科学), 2020, 40(3): 88–94 . DOI: 10.11929/j.swfu.201908022
Minjie Duan, Xinyu Li, Songting Zhao, Rui Xu, Yuerong Wang, Qian Wang. Study on Changes of Hyperspectral Characteristics of Plant Leaves Under Different Dust Retention Conditions[J]. Journal of Southwest Forestry University, 2020, 40(3): 88-94. DOI: 10.11929/j.swfu.201908022
Citation: Minjie Duan, Xinyu Li, Songting Zhao, Rui Xu, Yuerong Wang, Qian Wang. Study on Changes of Hyperspectral Characteristics of Plant Leaves Under Different Dust Retention Conditions[J]. Journal of Southwest Forestry University, 2020, 40(3): 88-94. DOI: 10.11929/j.swfu.201908022

不同滞尘环境下植物叶片高光谱特征变化研究

Study on Changes of Hyperspectral Characteristics of Plant Leaves Under Different Dust Retention Conditions

  • 摘要: 利用FieldSpec Pro野外光谱仪和电子分析天平对北京玉渊潭公园周边不同环境下的大叶黄杨叶片滞尘量、叶片除尘前后的光谱反射特征进行了测定与分析,通过微分变换,引入多组变量,研究了滞尘对大叶黄杨叶片光谱特征的影响,并建立了光谱特征参数与园林植物滞尘量的关系模型。结果表明:不同环境下大叶黄杨单位叶面积滞尘量表现为公园周边主干道>公园外围防护绿带>公园园内。叶片除尘前后光谱反射率存在明显差异,除尘后的光谱反射率值整体大于除尘前,其中在750~1 350 nm和1 500~1 850 nm波段差异显著。除尘前后三边位置没有发生明显改变,而三边幅值和三边面积均表现出除尘后大于除尘前,叶片滞尘对黄边、蓝边幅值以及黄边、蓝边面积影响不明显,但对红边幅值与红边面积影响较大。叶片滞尘对光谱特征参数均有影响且表现出一定规律性,其中利用红边面积构建的滞尘抛物线模型最优,估算精度及验证系数均相对较高,为基于高光谱技术估测植物叶片滞尘量提供了重要参考。

     

    Abstract: Spectrograph and electronic analytical balance were used to measure and analyze the dust retention and spectral reflection characteristics of the leaves of Euonymus japonicus before and after dust removal in Yuyuantan Park and the surrounding main roads. By analyzing multiple variables, the effect of dust on the spectral characteristics of E. japonicus leaves was studied and the relation model between spectral characteristic parameters and dust retention of plants was established. The results show that the dust retention per unit leaf area of E. japonicus shows a trend of that the main road around the park > that of the peripheral protection green belt > that of in the park. The spectral characteristics of leaves before and after dust removal generally show the trend of clean leaves > dust leaves and are significantly different in the bands of 750−1 350 nm and 1 500−1 850 nm. The three-sided positions have not changed after dust removal, while the three-sided amplitude and three-sided area all show that the trend of clean leaves > dust leaves. The influence of dust on the yellow-sided amplitude and area, the blue-sided amplitude and area are not obvious, but the influence on the red-sided amplitude and area are relatively significant. The leaf dust retention has an effect on spectral characteristic parameters and with certain regularity. Among them, the dust retention linear model constructed by using red-sided area has the highest estimation accuracy and reaches a significant level, which provides an important reference and data support for the estimation of dust retention of plant leaves based on hyperspectral technology.

     

/

返回文章
返回