how to extract data from table image?
8 views (last 30 days)
Show older comments
4 Comments
Walter Roberson
on 9 Jan 2019
Train the ocr on that character set.
The images are pretty low quality.
Benjamin Ehrlich
on 18 Aug 2021
I have experienced the OCR having a really rough time with vertical and horizontal lines...
Answers (1)
Image Analyst
on 19 Aug 2021
You need to have at least 20 lines in the numbers for ocr() to work -- it says so in the help. Also, you can get rid of lines by thresholding and removing the largest blob, but you might want to use them to extract each rectangle one at a time so you can do ocr on just that one rectangle. If you still need help, attach a higher resolution image
2 Comments
Abdelrahman
on 11 Mar 2023
Edited: Abdelrahman
on 11 Mar 2023
Can you elaborate more how to get rid of the lines please?
Image Analyst
on 11 Mar 2023
If the lines are connected to the edge of the image you can do
mask = grayImage > somethreshold;
mask = imclearborder(mask);
otherwise, the grid looks like it would be the biggest blob in the mask you can use bwareafilt
mask = mask & ~bwareafilt(mask, 1);
See Also
Categories
Find more on Recognition, Object Detection, and Semantic Segmentation in Help Center and File Exchange
Products
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!