模糊技术与神经网络的混合算法在运动目标的识别与跟踪中的应用  

The Application of Combination of Fuzzy Algorithms and Neural Network in Tracking Moving Object

作　　者：姚剑[1] 刘其真[1] 张斌 何永保[1] 林世雄[2] 刘军[2] 

机构地区：[1]复旦大学计算机科学系,上海200433 [2]上海航天局第八设计部,上海200233

出　　处：《计算机工程与应用》2000年第1期62-64,共3页Computer Engineering and Applications

摘　　要：图象跟踪是运动图象处理中的关键环节，但是如何在强噪声下实现快速有效的图象跟踪却一直是难点。该文提出了一种模糊技术与神经网络结合的算法，它充分吸收了模糊算法与神经网络的优点，在近距离目标下以神经网络识别的静态可能性为主，而在远距离目标时以模糊算法识别的动态可能性为主，使其不仅在目标近即大时有很高的识别率，而且更令人鼓舞的是在目标远即几乎为点目标时也有很高的识别率。另外，该文还提出了计算各个目标优先级的算法。最后，以红外序列图象为例表明该文提出的算法的高识别率。Image tracking is a key part of moving image processing but it is difficult to achieve fast and efficient image tracking in a highly noisy background. We present an algorithm combining fuzzy system with neural network which judges near targets mostly by the static possibility computed by neural network while judges far targets mostly by the dynamic possibility computed by fuzzy algorithms. It takes full advantage of fuzzy system and neural network and not only has very high recognition ratio of far targets hut also has high recognition ratio of near targets. By the way, we present an algorithm to compute the priority of different target. At last we give a series of infra images to test the highly recognized ratio using our algorithms.

关 键 词：运动目标跟踪 模糊算法 神经网络 图象处理 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]