【摘要】 SmartRoot是基于开放源ImageJ的免费根系分析软件,通过软件分析根系照片可得到细根长度、表面积、体积和平均根直径等指标。该文以橡胶树(Hevea brasiliensis)细根为例,探索利用SmartRoot分析细根不同形态特征的可能性。用不同拍照距离与分析方式的对比确定橡胶树细根适合的研究方法,用WinRhizo软件分析结果验证结果的准确性。结果表明,不同拍摄高度下对应的细根长度、表面积、体积和平均根直径等指标的标准均方根误差(NRMSE)分别为2.87%、15.73%、32.38%和16.88%;手动分析与半自动分析相同照片细根长度、表面积、体积和平均根直径等指标的NRMSE分别为3.06%、21.00%、40.96%和11.64%;用WinRhizo软件验证SmartRoot分析细根长度、表面积、体积和平均根直径等指标结果NRMSE分别为5.31%、9.37%、9.61%和5.77%。本研究表明,在适当方法下,拍照法结合SmartRoot软件分析测量橡胶树细根细根长度、表面积、体积和平均根直径等指标有较高的准确度,可用于橡胶树细根指标的分析。更多还原

【Abstract】 Aims Our objective was to examine the possibility of using SmartRoot,free software based on ImageJ,for the measurement of fine root length,surface area,volume and diameter of Hevea brasiliensis.Methods Photos were taken at heights of 15 and 40 cm from the fine root samples before the samples were subjected to comparison for accuracy of measurement of fine root length,surface area,volume and diameter of H.brasiliensis.The same images were analyzed manually and semi-automatically as well.Then the results obtained from SmartRoot were compared with those from WinRhizo,an expensive,popular,professional software for root analysis.Important findings The average normal root mean square errors(NRMSE) between 15 and 40 cm height for fine root length,surface area,volume and diameter were 2.87%,15.73%,32.38% and 16.88%,respectively.In addition,the NRMSE between manual and semi-automatic analysis for the same morphological parameters were 3.06%,21.00%,40.96% and 11.64%,respectively.Furthermore the NRMSE between SmartRoot and WinRhizo on fine root length,surface area,volume and diameter were 5.31%,9.37%,9.61% and 5.77%,respectively,signifying a high consistency between the two software systems.The results indicated that it is feasible to employ the SmartRoot system for determination of fine-root morphological parameters of H.brasiliensis due to its satisfactory accuracy,reliability,easy operation and time-saving advantages.For better accuracy,photos are recommended to be taken at the height of 15 cm and analyzed with the semi-automatic method.更多还原