数码照相法估算植被盖度

Vegetation coverage assessment by digital photos

植被盖度是表征生态系统特征的重要参数之一,其准确测度对于分析土壤侵蚀强度、生态环境现状具有重要意义。该文选择黄土高原退耕还林还草区野外测量的25个典型样方,探讨了数码照相法估算植被盖度的模型和方法,提出了中心投影引起的边缘畸变问题和植被信息识别问题的有效解决方案。前者采用控制点几何校正的方法解决,后者基于目视判读、统计汇总红(R)、绿(G)、蓝(B)三基色的DN(digital number)值分布规律,采用编程方式解决。研究结果表明,植被信息的三基色组成分别符合DNG>DNR>DNB、DNG>DNB>DNR和DNB>DNR>DNG3种规律,非植被信息的三基色构成遵循DNR>DNG>DNB的规律,且采用上述规律,编程得到草地样方的植被盖度精度较高。可见,该文提出的估算植被盖度的数码照相法可移植到数码相机中,具有明显的应用价值。

Vegetation coverage is one of the important parameters denoting ecosystem characteristics, so it is the intensity of soil erosion and identify the condition of environment for acquiring vegetation coverage of the study area. The paper chooses twenty-five key meadow sample plots in the Loess Plateau, then discusses the estimation model and estimation method of vegetation coverage by the pictures taken by the digital camera, and puts forward the effective solution for edge distortion problem caused by the central projection and vegetation information recognition problem. The former problem adopted geometry correction method by the control points, and the latter problem was solved by visual interpretation to take statistics for the distributing laws of three primary color DN （digital number） values of red（R）, green（G） and blue （B）. Finally, it is concluded that the vegetation spectral characteristics obey three laws of DNG 〉 DNR 〉 DNB , DNG 〉 DNB 〉 DNR and DNB 〉 DNR 〉 DNG , the non-vegetation spectral characteristics obey the law of DNR 〉 DNG 〉 DNB. And vegetation coverage values estimated by digital photo method have high precision. So the program of the vegetation coverage estimation by digital photos can be transplanted to the digital camera, and it has obvious application values.