Write characters into a training file.
The operator write_ocr_trainf_image is used to prepare the training with the operator trainf_ocr_class_box. Hereby regions, representing characters, including their gray values (region and pixel) and the corresponding class name will be written into a file. An arbitrary number of regions within one image is supported. For each character (region) in Character the corresponding class name must be specified in Class. The file name can be chosen at will. In contrast to write_ocr_trainf one image per character is passed. The domain of this image defines the pixels which belong to the character. The file format can be defined by the parameter 'ocr_trainf_version' of the operator set_system.
|
Character (input_object) |
image(-array) -> object : byte |
| Characters to be trained. | |
|
Class (input_control) |
string(-array) -> string |
| Class (name) of the characters. | |
|
FileName (input_control) |
filename -> string |
| Name of the training file. | |
| Default value: ''train_ocr'' | |
If the parameters are correct, the operator write_ocr_trainf_image returns the value 2 (H_MSG_TRUE). Otherwise an exception will be raised.
write_ocr_trainf_image is reentrant and processed without parallelization.
threshold, connection, create_ocr_class_box, read_ocr
trainf_ocr_class_box, info_ocr_class_box, write_ocr, do_ocr_multi, do_ocr_single
write_ocr_trainf, append_ocr_trainf
Optical character recognition
