python - Retrieve character' images from text line image -


i trying extract character' images text line image, can feed these images k-nearest neighbor classification (i building own ocr system).

i have retrieve text line image, , wonder how should proceed extract characters.

my first attempt use horizontal projection cut images (from binary image): horizontal projection

my second attempt retrieve contours connected components, , tread them separated characters. attempt results, example letter 'i' cannot retrieved because of 2 disconnected contours.

both these attempt failed when 2 characters too close (or collapsed) on each other.

do have suggestions? i'm trying way combine 2 of them still unsuccessful.

note: learning purposes. that's why don't want use existing solutions, except using opencv normal image processing. k-nearest neighbor mandatory, since it's main part of ocr system.


Comments

Popular posts from this blog

php - How to add and update images or image url in Volusion using Volusion API -

javascript - jQuery UI Splitter/Resizable for unlimited amount of columns -

javascript - IE9 error '$'is not defined -