Defects in the specimen of missile engine shell was detected by thermal wave image technology in this paper. In order to gain intuitionistic and accurate space structure image of the detected object, subtracting background and high-frequency emphasized filtering method were used to enhance the image quality. Then the defect was segmented from the background using particle swarm fuzzy clustering algorithm, while the defect size and depth were identified quantitatively. On this basis, 3D reconstruction of the defect by thermal wave image was recognized by Volume Rendering method. The results show that the precision of the defect quantitative identification is higher, and 3D reconstruction result is well, which help us to observe the location and size of defects intuitionisticly.