how to recognize the blank between two English words

[ghost]

How to recognize blank between two English words? for my current model, if I input one English sentence then the output will concatenate all the English words together. for example: inputed image: recognized result:

A------l-ll-t-h--e--r-e-c--o--g-n-i-tiio---n--a-c-cc--ur-a--c-i-e-s---o--n--t-h---e => Alltherecognitionaccuraciesonthe

So how to recognize the blank between two English words?

Thank you so much.

[YacobBY]

In the CRAFT project you need to set link_threshold and low_text higher so that it separates the words before you pass them as images to the OCR. CRAFT is responsible for cleanly separating text, this project just reads the images without knowing what a space is.

[ghost]

@YacobBY thanks for the reply.

[rexlow]

@YacobBY Yes, the CRAFT project does exactly that. But afterward, how do we connect back the words that belong in the same sentence?

[fingoldo]

@YacobBY Yes, the CRAFT project does exactly that. But afterward, how do we connect back the words that belong in the same sentence?

probably by computing proximities of bounding boxes

[AhmadZobairSurosh]

@YacobBY

Hi Did you find a solution for this problem? I mean connecting back the words that belong in the same sentence. In my case its numbers separated by .

Thanks in advance

[YacobBY]

@AhmadZobairSurosh There is no real way to do it with deep-learning, probably just check whether the word next to it is on the same vertical pixel height within a margin and not more than a certain amount of horizontal pixels away.