Read training characters from files and convert to images.
read_ocr_trainf reads all characters from the specified file names and converts them into images. The domain is defined according to the foreground of the characters (as specified in write_ocr_trainf). The names of the characters are returned in CharacterNames. If more than one file name is given the files are processed in the order the file names.
|
Characters (output_object) |
image-array -> object : byte |
| Images read from file. | |
|
TrainFileNames (input_control) |
filename.named(-array) -> string |
| Names of the training files. | |
| Default value: '' | |
|
CharacterNames (output_control) |
string-array -> string |
| Names of the read characters. | |
If the parameter values are correct the operator read_ocr_trainf returns the value 2 (H_MSG_TRUE). Otherwise an exception handling is raised.
read_ocr_trainf is reentrant and processed without parallelization.
disp_image, select_obj, zoom_image_size
Optical character recognition
