Skip to content

MCV-2019-M1-Project/Team4

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Team 4

Members: Sergio Casas Pastor, Sanket Biswas, Josep Brugués i Pujolràs

Week 3

The tasks for the second week include:

  • Filter noise with linear or non-linear filters.

  • On denoised images, detect box with overlapping text, and apply OCR to get the text. Test query system using QSD1-W2 using only text.

  • Implement texture descriptors, and test them using QSD1-W2 using only texture descriptors.

  • Combine descriptors (text + color, text + texture, texture + color, text + color + texture) and test the retrieval on QSD1-W3.

  • Repeat the previous analysis with QSD2-W3 (remove noise, remove background, find 1 or 2 paintings and return the correspondences.

Slides for Week 3 are available here: Slides

How to run

main.py <querySetPath> <backgroundRemoval> <textRemoval> <textRemovalMethod> <k> 

The querySetPath parameter indicates the location of the query set images.

The backgroundRemoval parameter indicates if background has to be removed from the query set images.

  • True: background removal
  • False: no background removal.

The textRemoval parameter indicates if the text has to be removed from the query set images.

  • True: text removal
  • False: no text removal.

The textRemovalMethod parameter indicates the method used for text detection:

  • 1 : text detection based on color segmentation
  • 2 : text detection based on morphology operations
  • 3 : text detection based on EAST opencv neural network

The k parameter is used to indicate the number of top results that need to be saved.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages