This is the implementation of the paper A Technique to Detect Books from Library Bookshelf Image by Mohammad Imrul Jubair and Prianka Banik. The paper was published in IEEE 9th International Conference on Computational Cybernetics, July 8-10, 2013, Tihany, Hungary.
link: https://ieeexplore.ieee.org/document/6617619/
The code is written by Mohammad Imrul Jubair (author of the paper)
- open and run the book_recogn.m
- input images must be located inside INPUT folder.
- OUTPUT folder contains all the results. There are sub folders - a. SHELF: contains extracted shelves. b. DETECT: line detected as book boundary (shown as yellow lines) for each shelves. c. BOOKS: extracted books. There is a naming format for each images: shelfNumber_bookNumber.jpg. For example, 2_7.jpg means this is extracted from shelf no. 2, book number 7.
Note: Some sample input images are collected via google image searching. Another implementation can be found here: https://www.mathworks.com/matlabcentral/fileexchange/55134-a-technique-to-seperate-books-from-a-shelf-and-read-characters-from-it