中文摘要：

胸径测量是评价立地质量和林木生长状况的重要依据之一。利用普通数码相机对待测样地进行上下方向的任意摄影，并在林地中测量任意一段物体长度以便在解算中反推摄影基线，恢复摄影区域的真实空间比例关系，实现在图像上测量样地样木胸径和树心坐标的位置。该研究以摄影测量的理论为基础，以同名光线和摄影基线共面为条件，针对林业摄影测量的局限性，建立以垂直地面方向为主方向的摄影测量方式，实验分析验证测量算法的有效性，以MicrosoftVisualStudio为开发平台，编制地面立体摄影测量系统软件，对5组样地25株立木进行测量计算，实验表明，采用普通数码相机测量样木胸径的精度高，25棵样木的平均绝对误差为0．29cm，平均相对误差为1．99％，符合国家森林资源连续清查中关于胸径测量的精度要求，该方法在森林资源调查中具有实际的应用前景。

英文摘要：

The measurement of tree diameter at breast height （DBH） is an important parameter for evaluating site quality and tree growth status. An ordinary digital camera was used to randomly shoot plot that was measured in an up and down direction. And the length of object should be measured so as to reversely deduct the position of photographic baseline and recovery the true spatial scale of photography area, achieving the aim for measuring the DBH of sample trees in the plot and the coordinate of the tree＇s center. This research was based on theory of photogrammetry and carried out with the coplane of rays from same point and photographic as the conditions. Aiming at the limitation of photogrammetry, this research established the photogrammetry mode mainly in a vertical ground direction. The validity of the proposed algorithm was analyzed and verified in the experiment. Software of ground stereo photogrammetry system was developed, taking Microsoft Visual Studio as development platform. Twenty-five standing trees in five plots were measured and calculated. Experiment results demonstrated that the precision of DBH measurement using ordinary digital camera was high, and the experimental data were in good agreement with the actual, with the average absolute error of 0. 29 cm, and the average relative error of 1.99%, which were in line with the tree diameter at breast height measurement accuracy of national forest resource continuous investigation. With the constant increase of field working cost, people will pay more attention on the utilization of portable instruments in measurement. Therefore, the method has practical application prospect in forest resource survey.