Read an OCR classifier from a file.
read_ocr_class_mlp reads an OCR classifier that has been stored with write_ocr_class_mlp. Since the training of an OCR classifier can consume a relatively long time, the classifier is typically trained in an offline process and written to a file with write_ocr_class_mlp. In the online process the classifier is read with read_ocr_class_mlp and subsequently used for classification with do_ocr_single_class_mlp or do_ocr_multi_class_mlp.
HALCON provides a number of pretrained OCR classifiers (see Quick Guide, chapter 'OCR', section 'Pretrained OCR Fonts'). These pretrained OCR classifiers make it possible to read a wide variety of different fonts without the need to train an OCR classifier. Note that the pretrained OCR classifiers were trained with symbols that are printed dark on light.
|
FileName (input_control) |
filename -> string |
| File name. | |
| Suggested values: 'Document_0-9A-Z.fnt', 'Document_0-9.fnt', 'Document.fnt', 'DotPrint_0-9A-Z.fnt', 'DotPrint_0-9.fnt', 'DotPrint_0-9+.fnt', 'DotPrint.fnt', 'HandWritten_0-9.fnt', 'Industrial_0-9A-Z.fnt', 'Industrial_0-9.fnt', 'Industrial_0-9+.fnt', 'Industrial.fnt', 'MICR.fnt', 'OCRA_0-9A-Z.fnt', 'OCRA_0-9.fnt', 'OCRA.fnt', 'OCRB_0-9A-Z.fnt', 'OCRB_0-9.fnt', 'OCRB.fnt', 'Pharma_0-9A-Z.fnt', 'Pharma_0-9.fnt', 'Pharma_0-9+.fnt', 'Pharma.fnt' | |
|
OCRHandle (output_control) |
ocr_mlp -> integer |
| Handle of the OCR classifier. | |
If the parameters are valid, the operator read_ocr_class_mlp returns the value 2 (H_MSG_TRUE). If necessary an exception handling is raised.
read_ocr_class_mlp is processed completely exclusively without parallelization.
do_ocr_single_class_mlp, do_ocr_multi_class_mlp
create_ocr_class_mlp, write_ocr_class_mlp, read_class_mlp, write_class_mlp
OCR/OCV
