Detection of micro damages in flat components on-site has a significant sense for improving the safety of the equipment. Based on the theory of the surface acoustic wave (SAW) propagation laws in flat components, the micro damage detection in the flat component has been researched. Using wavelet analysis technology and inversed spectrum technology, the microdamages' feature parameters can be extracted out accurately. Utilizing the feature parameters got by every transducer in a transducer arrays, the micro-damages' image can be reconstructed and the micro-damages' location, outer geometric configuration, and damage level can be showed clearly.