Skip to content

A Panoramic Image stitching implementation using classical and deep learning method

Notifications You must be signed in to change notification settings

naitri/panorama-stitching

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Panorama Image Stitching

This repository implements two approaches for panorama image stitching. The first is traditional method of image stitching using corner detection, Adaptive non-maximal suppression, feature descriptor, matching and RANSAC. The another is Deep Learning Approach where supervised and unsupervised approach is explored.(HomographyNet)

Traditional Approach

Original Undistorted Undistorted

Pano

The path to the folder of images can be provided as mentioned below.

python Wrapper.py --Folder $PATH_TO_DATA

Deep Learning Approach

  • The Wrapper.py simply shows an example of generating a pair of images according to the Data Generation technique presented in the Supervised approach paper.
  • To train the supervised model run python Train.py --ModelType=sup
  • To train the unsupervised model run python Train.py --ModelType=unsup
  • To test the supervised model run python Test.py --ModelType=sup after having trained it.
  • To test the unsupervised model run python Test.py --ModelType=unsup after having trained it.
  • All scripts assume that there is a Data folder, in which MS COCO images are stored according to the given filesystem scheme.