Condition monitoring of wooden railway sleepers applications are generally carried out by visual inspection and if necessary some impact acoustic examination is carried out intuitively by skilled personnel. In this work, a pattern recognition solution has been proposed to automate the process for the achievement of robust results. The study presents a comparison of several pattern recognition techniques together with various nonstationary feature extraction techniques for classification of impact acoustic emissions. Pattern classifiers such as multilayer perceptron, learning cector quantization and gaussian mixture models, are combined with nonstationary feature extraction techniques such as Short Time Fourier Transform, Continuous Wavelet Transform, Discrete Wavelet Transform and Wigner-Ville Distribution. Due to the presence of several different feature extraction and classification technqies, data fusion has been investigated. Data fusion in the current case has mainly been investigated on two levels, feature level and classifier level respectively. Fusion at the feature level demonstrated best results with an overall accuracy of 82% when compared to the human operator.