通过对7种柳属植物的350枚叶片进行实验处理，测量所需的数字化植物特征，在单一指标判别分析中，仅有“长宽比”作为识别这7种植物的最有效的单一指标；在双重指标的判别分析中，“最宽处”ד叶脉值”、“最宽处”ד短弧距”、“最宽处”ד长弧距”、“周长比”ד膜周长比”、“短弧距”ד上外距”、“长弧距”ד上外距”这6组双重指标对7种植物的识别正确率均达到100％，同为最优双重指标。试验结果表明，上述6组最优双重指标可以完全识别出这7种植物。因此，基于数字化的叶片图像特征的信息识别物种是可行的，为今后植物物种识别研究奠定了基础。%We studied an effective method for recognizing digital indicators of seven Salix species.In single index discriminant analy-sis, only“aspect ratio” is the most effective index for the identification of these seven species .In double index discriminant analysis, the recognition correct rates of seven Salix species of six sets of double index , “width”דvein value”, “the width”דshort arc distance”,“width”דlong arc distance”,“perimeter”דmembrane perimeter”,“short arc distance”דon the pitch”, and “long arc distance”דon the pitch” are up to 100%, and six sets are all the optimal dual index . These six sets of double index can be used to completely identify seven Salix species.
展开▼
