You might be able to reduce error beyond 2% by eliminating extra pixels in the margins (like the one you suspected of causing the error). I imagine you could achieve that through a script that sums up pixel values in a single axis to detect text edges and clear everything outside the text part of the image. You may further improve it to even detect and clear artifacts in spacing between text lines, further lowering your margin for error.
Hello thank you for the video. Is there a way to get the image preprocessed by the tesseract algorithm? When running tesseract in cmd I can get it by setting tessedit_write_images = 1, but in python I couldn't find a way to get preprocessed image.
hello, i have very limited knowledge on how to use the command line and how to add things to my directories in my terminal on mac; could you please assist me in trying to achieve this haha, i am having a very hard time trying to get tesseract to work on my mac
@@python-programming Thanks, I just finished watching all the playlist. I will definitely try this preprocessing stuff. Anyway, I have to OCR some numbers from a sensor's screen, I was wondering if there are other method to automatically read that numbers. Maybe some deep learning stuff or similar things. Any suggestions? Thanks again
Your work is so underrated. Really thanks for helping us!❤
Thank you so much. I’ve gone through a lot of videos but your presentation and explanation is way superior.
I just saw this comment. Thanks so much!!
Thank you for the OCR series, Sir. Eagerly waiting for more OCR tutorials.
I have one more NER video and then 6 OCR videos. Late May and early June will be all OCR
@@python-programming I'm excited about your new videos.
Thanks!
About a year late, but this series helped me execute my project idea, thanks a lot!
No problem!!
You might be able to reduce error beyond 2% by eliminating extra pixels in the margins (like the one you suspected of causing the error). I imagine you could achieve that through a script that sums up pixel values in a single axis to detect text edges and clear everything outside the text part of the image. You may further improve it to even detect and clear artifacts in spacing between text lines, further lowering your margin for error.
Thank you so much. I really looking forward to implementing OCR from scratch without Tesseract any suggestions will be appreciated
thank you so much for these videos!
No problem! Glad you are enjoying them!
Hello thank you for the video. Is there a way to get the image preprocessed by the tesseract algorithm? When running tesseract in cmd I can get it by setting tessedit_write_images = 1, but in python I couldn't find a way to get preprocessed image.
I am not sure. If you wanted to do it programmatically, have you considered subprocess?
Can we use this for handwritten notes?
hello, i have very limited knowledge on how to use the command line and how to add things to my directories in my terminal on mac; could you please assist me in trying to achieve this haha, i am having a very hard time trying to get tesseract to work on my mac
Hi, great series. Helping me a lot.
Is there a way to improve furthermore the result of OCR's? Any other method or something?
Thanks.
The next few videos show you some methods but really it is a case by case basis depending on the quality of the scan and language of the text.
@@python-programming Thanks, I just finished watching all the playlist. I will definitely try this preprocessing stuff.
Anyway, I have to OCR some numbers from a sensor's screen, I was wondering if there are other method to automatically read that numbers. Maybe some deep learning stuff or similar things. Any suggestions?
Thanks again
Hi , i need OCR using template matching can you help me
?
how to custom train tesseract 5.4.0 pr 5.4.1 for nepali script??
Hey please help in installation of tesseract in ubuntu20.04
If you could help me I promise you that I will be a patreon of your channel.
Sorry. I have never installed it on Ubuntu before, but here is a link linuxhint.com/install-tesseract-ocr-linux/