We currently plan to make the dataset content publicly available, but we do not intend to open source any other code. In our paper, the 3D reconstruction is achieved by first calculating the left-eye disparity from stereo images captured by a binocular camera. The resulting disparity map is converted into a depth map using the camera's intrinsic parameters, and this depth map is then triangulated to generate point cloud data. Our dataset provides ground truth depth values to assess the accuracy of the 3D reconstruction. For detailed camera parameters, please refer to the paper.
We currently plan to make the dataset content publicly available, but we do not intend to open source any other code. In our paper, the 3D reconstruction is achieved by first calculating the left-eye disparity from stereo images captured by a binocular camera. The resulting disparity map is converted into a depth map using the camera's intrinsic parameters, and this depth map is then triangulated to generate point cloud data. Our dataset provides ground truth depth values to assess the accuracy of the 3D reconstruction. For detailed camera parameters, please refer to the paper.