A new approach to the retrieval of fringe pattern frequency information is introduced. In this approach, strain information is extracted from fringe patterns by using either short-time Fourier transforms or wavelets. Space-frequency representation of the fringe signals is introduced as a tool to get the space localization of the strains. The idea of spectral energy density maxima to define the instantaneous frequency (ridge points) is presented. Examples of application are given. Results show that this approach provides an excellent tool for pattern analysis.