Above is the picture, I want to segment character in TEST to 4 separated images, each contains T,E,S,T respectively and then use CNN to recognize each character. How many algorithms could help this? Could I use RNN with LSTM to do it? Because I prefer neural network but if there is any easier algorithm, please indicate. Thank you so much.
Asked
Active
Viewed 39 times
0
-
besides The answer here for more speed these are used too: [Interpretation of Horizontal and Vertical Summations of an Image](https://stackoverflow.com/a/39178204/2521214) and [Detecting space(bar) between words in a slanted font](https://stackoverflow.com/a/31086741/2521214) to detect font size and line configuration ... – Spektre Nov 16 '17 at 06:25
1 Answers
1
On English, all letters are contiguous areas, so you need to do binarization and select (big enough, and not too big) contiguous areas.

user31264
- 6,557
- 3
- 26
- 40
-
This is true for capital letters only, not for lowercase 'i' and 'j'. – m69's been on strike for years Nov 17 '17 at 00:06
-