利用航空数码影像的色彩特征估测森林郁闭度

Estimating the Canopy Density Through Color Features of Aerial Digital Images

采用高分辨率航空数码影像,结合色彩特征估测郁闭度。首先利用阴影的光学特性,在I、H、S空间对阴影区域亮度、饱和度和色调值进行匹配补偿,去除阴影。然后利用I1I2I3空间I2I3通道对绿色良好的敏感度和光照不变性,提取树冠并计算郁闭度,最后分析估测郁闭度的精度。结果表明,总体估测精度为99.89%以上,郁闭度区间的复层针叶林估测精度为87.18%,其他林分垂直结构和树种类型条件下的估测精度不低于95%。

The forest canopy density was estimated from the high resolution aerial digital images based on their color features.Firstly,the shades in the images were removed based on their optical features by adjusting the brightness,saturation levels and tonal values in space I,Hand S.Secondly,by taking advantage of channel I2 I3（in space I1 I2 I3）＇s high sensitivity to green color and illumination invariance,the data of the canopy were extracted and the canopy density was calculated.Thirdly,the accuracy rates of the estimated canopy density was analyzed.The results showed that the general accuracy rate could be as high as more than 99.89%.Except that the accuracy rate of stratified coniferous forest in the canopy density ranging from 0.20to 0.44was 87.18%,other estimated accuracy rates under the condition of the forest variety in vertical structure and types were no less than 95%.The canopy density of forests in subdued topography could be accurately estimated by combining high resolution aerial digital images and color features.The paper provided a valid method to accurately estimate the canopy density.