A novel fuzzy neural network (FNN) model for invariant pattern recognition is presented that combines fuzzy set reasoning and artificial neural network techniques. The presented FNN consists of three blocks: fuzzifier, fuzzy perceptron, and defuzzifier. It fuzzifies the input patterns and trains the interconnection weights according to membership functions instead of traditional binary values. The proposed FNN has been applied to 2-D binary-image pattern recognition under shift and some other types of distortions. In comparison with the classical multilayer perceptron, the FNN possesses a higher recognition rate and is more robust to input distortions.