Usually when OCR an table of content the columns are separated by a large space, so the outputs are not properly order. For example, for an table like this:
The output would be:
The Rank FunctionPermutations of AtomsPure Set Theory and Axiom System ZF3.53.63.7I'd like it to be:
3.5 The Rank Function\1123.6 Permutations of Atoms\1163.7 Pure Set Theory and Axiom System ZF\118But different TOCs has different the output patterns, so there is no way to build a regex script to automatically fix every book. The best approach is to fix it at the first place. But how?
- I haven't explored that tool much so I have no ideaOoker– Ooker2024-08-01 17:49:11 +00:00CommentedAug 1, 2024 at 17:49
2 Answers2
Define what is:"fix it at the first place".
If you want to fix wrong output from OCR analysis, a simple solution on an infinite set of TOCs you will never make.You will never apply all variations. You would have to create a machine learning algorithm that would analyze each TOC variant.
Or count substrings of the same characteristics (in simple TOC).
Chapter numberChapter numberChapter numberChapter numberChapter number...= 5
Chapter titleChapter titleChapter titleChapter titleChapter title...= 5
If you want to fix OCR analysis, it's a good to answer:What OCR tool do you use?
For example, inTesseract you can set, that text is processed by rows instead of columns.
- This sounds promising. What GUI do you use to make that screenshot? I've checked the3rd party projects but don't know which oneOoker– Ooker2018-07-10 05:21:08 +00:00CommentedJul 10, 2018 at 5:21
- I've opened a question in Software Recommendation:What OCR tools can recognize text by rows instead of columns?Ooker– Ooker2018-07-10 06:00:28 +00:00CommentedJul 10, 2018 at 6:00
- @Ooker : I think most of the set of OCR tools have the ability to set how the page should be analyzed. The picture is my schematic illustration. But this borders also shows FineReader and the advanced GUI for TesseractgImageReader (GitHub). But in automatic analysis you do not need a GUI if you do not want to manually edit something IMHO. Because you can not use automatic batch processing over a large amount of TOC if you want to check and edit each TOC.Stilgar Dragonclaw– Stilgar Dragonclaw2018-07-10 18:12:20 +00:00CommentedJul 10, 2018 at 18:12
- @Ooker your question on OCR tools recommendation is removed. Did you move it to somewhere else?raf– raf2024-08-02 04:12:44 +00:00CommentedAug 2, 2024 at 4:12
- 1@raf it was deleted by the roomba for being inactive.I reask it. Hope it lives longerOoker– Ooker2024-08-02 14:35:30 +00:00CommentedAug 2, 2024 at 14:35
Not really answer the question, but some books in Google Books have TOC:
Explore related questions
See similar questions with these tags.