Abstract: Aiming at the problems of being non intuitive, of high false detection rate and high undetected rate of traditional artificial detection results of aero-engine blades, a 3-D imaging method based on blade ultrasonic point cloud is proposed. The display of ultrasonic point cloud 3-D reconstruction model is completed through the research of ultrasonic C-scan imaging principle, space surface display technology, 3-D shading model and visualization technology. The results show that the reconstruction algorithm is effective by using Delaunay triangulation algorithm to reconstruct the blade point cloud. A 3-D shading model based on OpenGL is established, and the general shading mapping space is changed from RGB color space to CIELAB color space, which makes the display of defect area more reasonable. The analysis and display of ultrasonic point cloud are accomplished by using the method of mixed programming of C#.NET and MATLAB. The specific location of blade defects in the overall model can be visually shown by using the 3-D visualization method. It enables the inspectors to observe the 3-D model in an all-round way, which facilitates the rapid location and detection of defect areas, and greatly improves the efficiency and accuracy of detection.