I have extracted 34 unique words(provinces names) from the number plate of cars. The examples images are attached. Now how can I classify the words? Could you please give me some code, comments, and recommendations?

Question

Javid Hamdard 2021-3-22

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/780112-i-have-extracted-34-unique-words-provinces-names-from-the-number-plate-of-cars-the-examples-images

回答： Image Analyst 2021-3-22

采纳的回答： Image Analyst

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Image Analyst 2021-3-22

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/780112-i-have-extracted-34-unique-words-provinces-names-from-the-number-plate-of-cars-the-examples-images#answer_654857

You can try a deep learning approach. Follow the example that uses NIST digits and adapt it.

Another more traditional approach may be just computing the area fraction of the images. It could be that every state has a different number of white pixels in it. Or just call imabsdiff() of the test image with all reference images. The one with the lease differences will be the state that it is.