Aircraft image recognition is an important sub-problem of photo-interpretation, which continues to be a major application of domain of image understanding techniques. This paper describes a new approach to aircraft image recognition system. This approach combines complex moments with complex neural-network. For aircraft recognition using complex moments are a recent one of the most useful features that can be extracted from an image because they can be invariant to translation, rotation, and scaling of the object. A neural network systems are adept at many pattern recognition tasks which require the ability to match large amount of input information simultaneously and then generate categorical or generalized output. Neural network systems posses these capabilities as well as the ability to learn and build unique structures specific to a particular problem, so they are especially useful in pattern recognition. This paper presents complex neural classifier, which uses complex moment function as inputs. This approach differs significantly from many recent aircraft recognition system, which emphasize the geometric features of objects in the scene, experimental results are given to illustrate the efficacy of this approach.