/CV-Project

Primary LanguagePython

CV Project -- Stereo Image for Depth Map

TODO:

  • Find Essential matrix or Fundamental matrix (in func.py) (version 0.2)
  • Find translation vector (T) (in func.py) (version 0.4)
  • Find rotation matrix (R) (in func.py) (version 0.6) (using openCV to solve)
  • Rectified two images (in rectified.py) (version 0.8) (using openCV to solve)
  • Get the Disparity Map (in correspondence.py) (version 0.9)
  • Get the Depth Map (in func.py) (version 1.0)
  • The unstable reason because for using RANSAC to get F Change the feature extractor from SIFT to ORB (solved!)
  • Paper1 implement : Depth estimation from stereo images using sparsity
  • Paper2 implement : Adaptive support-weight approach for correspondence search

Dataset:

Result

  • Original
    Test

  • Paper1
    Test

  • Paper2
    Test

Reference:

  1. Get T and R from Essential Matrix.
  1. Similar work