how to use unicode with knearest::train for arabic character
I want to do something like that but when I do that the result is ASCII code, not Unicode
cv::Mat_<wchar_t> trainFeatures(40, 1);
trainFeatures << 'ب', 'ت', 'ث', 'ج', 'ح', 'خ', 'د', 'ذ', 'ر', 'ز', 'س', 'ش', 'ص', 'ض', 'ط', 'ظ;
cout << trainFeatures;
train data type must be CV_32F : no problem with unicode
already i use it but when i cout the mat give me ASCII values not Unicode values
@walid, hamdy, you might need to explain, what you're trying to do here, and what you expect to achieve.
in general, character recognition works on image features, not plain text (whatever that is encoded in)
cv::Mat does not support wchar_t at all. it probably only compiles, because your compiler assumes, it's the same as unsigned short
again, you will need numbers for both train data and labels, not (unicode) characters.