Multi-view 3D Reconstruction of the Flame Surface Based on Deep Learning
To address the background noise in the 3D reconstruction of the flame surface,an IM-MVSNet network based on the MVSNet multi-view 3D reconstruction network was proposed for reconstructing the flame surface of laminar flow flames.The network obtained high-quality segmented images by image segmentation of the reference frames and neighboring frames of the input sampled images to remove the background noise during sampling,and then reconstructed the multi-view images in 3D to build a 3D point cloud of the laminar flame surface,and then obtained the reconstructed laminar flame surface.The reconstruction results of the flame surface of different reconstruction models show that the 3D reconstruc-tion network proposed in this paper can effectively reduce the point cloud noise of the reconstructed flame surface,improve the reconstruction accuracy of the flame surface,and provide a new technical means for combustion research.
万方数据
维普
