用证据累积的方法识别不完整图象

RECOGNIZING THE IMPERFECT IMAGE BASING ON THE EVIDENCE ACCUMULATION METHOD

本文将专家系统中处理不精确信息时所采用的基于确定性理论的推理模型引入到对不完整图象的目标识别中来，提出了利用证据累积的方法对目标进行识别的观点，强调所有信息的综合作用和正反两方面证据对可信度的影响。

In this paper, the CF(Certain Factor) theory which is one of well--known reasoning modes used in research of expert system for dealing with the uncertain information is introduced into the processes of recognition of imperfect image and a new method is proposed which works on the principle of evidence accumulation. This method emphasizes the effect of fall-aspects information to avoid the error decision which may be caused by uncertainty of individual feature.