Abstract:Recently the depth estimation methods based on deep learning (DL) retains challenging to estimate high-precision in fringe projection structured light three-dimensional(3D) measurement with limited information from a single-frame fringe pattern. In this letter, we proposed a FDSUNet++ convolutional neural network, which consists of a UNet++ base model, an improved squeeze-and-excitation (ISE) block, a Fourier transform (FT) data preprocessing block, and a discrete wavelet transform (DWT) block. The proposed ISE block can improve the ability of feature extraction and the designed FT data preprocessing block preserves the key features of the fringe pattern by Fourier transform. The introduced DWT block reduces the complexity and training cost of the model. By integrating these three blocks into the UNet++ network, it can better achieve depth estimation. Experimental results from two structured light datasets demonstrate that the pro-posed FDSUNet++ outperforms the state-of-art networks such as UNet, SUNet, R2U-Net, PCTNet and MSUNet++, achieving the best performance in both qualitative and quantitative evaluation.
