跳转至

Two-View Stereo Matching

约 85 个字

Two-View Stereo Matching

  • Task: Construct a dense 3D model from 2 images of a static scene

Pipeline:

  1. Calibrate cameras intrinsically and extrinsically (Lecture 3.1)
  2. Rectify images given the calibration
  3. Compute disparity map for reference image
  4. Remove outliers using consistency/occlusion test
  5. Obtain depth from disparity using camera calibration
  6. Construct 3D model, e.g., via volumetric fusing and meshing (Lecture 8.4)

  • Image rectification

    Goal:Correspondences search along horizontal scanlines (simplifies implementation)

    image-20240218144003587

    image-20240218144013565

    Block Matching

    Siamese Networks

    image-20240218144142889

    Loss function:

    \(L = max(0, m + s_- - s_+)\)

    Spatial regularization

    more smooth between pixels

    image-20240218144357978