konstantint
Cannot specify the character '<'
Hi, I am using the read_mrz function but after extraction names and surnames cannot fulfill the condition of splitting and replacing '<' just because mrz cannot be able to identify '<'. It is detecting 'K' or 'S' or 'G' or some other character instead of '<'. How can I improve it to get an accurate result?
Apart from algorithm improvements (which require some creative experimentation I couldn't give simple recipes for at the moment), your best bet is to get better quality images (or make sure the images are oriented correctly).
Konstantint, thanks for your quick response. Can you please tell me what would be the required quality/resolution/details of an image to get an accurate result? I mean is there a threshold or limitation of the image quality required for this algorithm? And one more thing, Can I modify your algorithm according to my requirements?
Take a look at images here. The number in the title is the known "recognition score" for that image, so images with 100 are generally easily readable.
Of course you are free to play with the algorithm and tune it to your needs, that's why this repository is here out in the open. Let me know if you find a good algorithmic solution to your problem that can be widely applicable.