Skip to content

Implementation of Joint Spectral Correspondence for matching the images with disparate appearance arising from factors like dramatic illumination (day vs. night), age (historic vs. new) and rendering style differences.

Notifications You must be signed in to change notification settings

chaitanya100100/Spectral-Analysis-for-Image-Matching

Repository files navigation

Joint-Spectral-Correspondence

Implementation of Joint Spectral Correspondence for matching the images with disparate appearance arising from factors like dramatic illumination (day vs. night), age (historic vs. new) and rendering style differences.

Project Description

  • Graph Spectral Analysis of image has been used for classic problems like image segmentation, classification, clustering. This project aims to match two images of the same place with dramatic lighting differences using joint spectral analysis of both image.

  • This method is based on this research paper accepted in CVPR-2013.

  • src/eigen_function has code to compute joint eigen functions of both images.

  • src/matching has code to match eigen functions of both images.

  • dataset contains two images for each place and homography between those two images.

  • results contains precomputed eigen functions in mat file for each image pair of dataset. Matching can be done by directly loading those mat files.

  • reports contains following :

    • Theory-basic of eigen spectrum of image : image-graph-spectrum-theory.pdf
    • Implementation details of this project : implementation.pdf
    • Generated eigen functions for some image pairs of dataset : eigen.pdf
    • Results after matching eigen functions of some image pairs : matching.pdf

Results

trevi02
stargarder vatican bdom

About

Implementation of Joint Spectral Correspondence for matching the images with disparate appearance arising from factors like dramatic illumination (day vs. night), age (historic vs. new) and rendering style differences.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published