This paper evaluates three different image representation techniques relative to fingerprint image modeling. The most popular image representation techniques such as Gabor filters, Discrete Fourier transform and Wavelet transform will be implemented and compared relative to fingerprint identification performance. The evaluation and comparison criteria using database of 80 persons with 12 fingerprint samples from each person. The multilayer perceptron neural network with back propagation training algorithm was used as a pattern classifier to evaluate the three mentioned image representation techniques. The obtained results showed that the average identification rate for the wavelet transform, Discrete Fourier transform and Gabor techniques are 97.9%, 93.9% and 88.1% respectively.