【摘要】 文中主要针对铁路红外图像的特点 ,在原有的自动目标识别算法上 ,将图像处理技术与人工智能相结合 ,提出一种基于知识的智能识别方法。采用先建立完善的结构知识库 ,再以知识库为基础 ,采用写生式策略来建立各种规则 ,并结合规则指导图像分割和特征提取 ,推断铁路的存在并指导铁路片断的连接。其基本思路可用于一般目标的识别。更多还原

【Abstract】 According to the characteristic of digital infrared image, a new method from original ATR methods is presented, which combines image processing and artificial intelligence. At first, a perfect frame knowledge base is built, and some rules are built using sketch\|style policy, and then image segment and lineal characteristic extraction are combined, at last, the existence and connection of railroad segments are induced. The basic processing route can be used for general ATR.更多还原