中文摘要：

针对橡胶种植适宜性评估，基于云理论、粗集理论和模糊神经网络理论，提出了一种适宜度评估模型。该模型将转化的样本数据进行粗集简约，通过模糊神经网络得出评价因子的隶属函数，计算评价等级。研究结果表明，此模型能够科学、快速、准确地分析出橡胶种植最适宜区、适宜区、次适宜区和不适宜区。

英文摘要：

The circular target had broad appication in close-range photogrammetry, the image of the circular target appeared as an ellipse because the target plane could not be completely parallel to the image plane. The deviation between circular target center and ellipse imaging center is inevitable if circular target is oversized. From collinear condition equation of photogrammetry, a deviation correction formula was derived based on least square ellipse fitting algorithm. The objective is to analyze distribution rule and the relations among deciations, so as to reduce the effect of deviation.