【摘要】 运用森林类型遥感目视识别的７０个样本，训练人工自组织神经树模型，然后对１０个“未知”样本进行预测。结果表明，该模型的识别、容错能力较强，综合了遥感图像专家目视判读与计算机自动识别的优点，使判读过程更加精确和简练，而且省工、省时、省经费。开拓了遥感识别地物的新途径更多还原

【Abstract】 A model of artificial self-organizing neural tree was trained by 70 sample books of forest type recognized by visual remote sensing recognization. The trained model was then evaluated by the recognition of 10 unknown samples. The results showed the abilities of recognition and fault tolerance of the model were strong. It combines the advantages of visual interpretation of remote sensing photoes with automatic computer recognition, which makes the recognition not only more accurate and concise, but also time, money and labour saving. It opens up a new way for remote sensing recognition of object.更多还原