Hello! I try to make a OCR with the Vision Assistant from LabView. If the characters "I" and "T" are merged, the Vision Assistant cuts the characters at the wrong place. (See picture below). I tried also the algorithm "shortest segment" but it didn't work correctly. NI Manuals say this: "The algorithm works in three steps. Attempt to divide the characters by applying multiple shortest path cuts. Choose the cuts that are closest to the max character width. Intelligently choose the cuts which segment a character correctly based on classification during reading." If my max character width is adapted to the width from the "T", the algorithim chooses the wrong cut to seperate the characters. If I decrease my max character width suitable for the "I", the "T" is not registered completely (see picture) . Does somebody has any idea to solve my problem?

LabVIEW

OCR: wrong segmentation if characters have different width

Hello!

I try to make a OCR with the Vision Assistant from LabView. If the characters "I" and "T" are merged, the Vision Assistant cuts the characters at the wrong place. (See picture below).

I tried also the algorithm "shortest segment" but it didn't work correctly.

NI Manuals say this:

"The algorithm works in three steps.

Attempt to divide the characters by applying multiple shortest path cuts.
Choose the cuts that are closest to the max character width.
Intelligently choose the cuts which segment a character correctly based on classification during reading."

If my max character width is adapted to the width from the "T", the algorithim chooses the wrong cut to seperate the characters. If I decrease my max character width suitable for the "I", the "T" is not registered completely (see picture)

Does somebody has any idea to solve my problem?