光路受限条件下基于轮廓的 PSP 图像重构方法

In the field of internal flow field with limited optical path, such as compressor cascade experiment, three-dimensional reconstruction of image is crucial for pressure measurement of Pressure Sensitive Paint (PSP). The key to the three-dimensional reconstruction process is to accurately obtain the projection matrix between the measured model and the camera, while the existing methods are limited by the number and distribution of feature points. In this paper, a Contour Matching-based Direct Linear Transformation (CM-DLT) is proposed. This method draws on the idea of iterative closest point algorithm, optimizes the projection matrix by matching the image contour and the re-projection contour of the 3D model, and improves the reconstruction accuracy of the 3D pressure. In the benchmark experiment, the effects of 3D contour point cloud dispersion, initial feature point noise, number and distribution on the CM-DLT algorithm are studied. Subsequently, the CM-DLT method was successfully applied to the compressor cascade PSP experiment, and the three-dimensional pressure distribution on the blade surface was obtained. The results demonstrate that reducing the dispersion of 3D contour point cloud is beneficial to improve the accuracy of CM-DLT algorithm. In addition, the CM-DLT method has strong anti-interference ability and can effectively realize the accurate reconstruction of three-dimensional pressure.