trainf_ocr_class_svmT_trainf_ocr_class_svmTrainfOcrClassSvmTrainfOcrClassSvmtrainf_ocr_class_svm (算子)

名称

trainf_ocr_class_svmT_trainf_ocr_class_svmTrainfOcrClassSvmTrainfOcrClassSvmtrainf_ocr_class_svm — 训练 OCR 分类器。

签名

trainf_ocr_class_svm( : : OCRHandle, TrainingFile, Epsilon, TrainMode : )

描述

trainf_ocr_class_svmtrainf_ocr_class_svmTrainfOcrClassSvmTrainfOcrClassSvmTrainfOcrClassSvmtrainf_ocr_class_svm trains the OCR classifier OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle with the training characters stored in the OCR training files given by TrainingFileTrainingFileTrainingFileTrainingFiletrainingFiletraining_file. The training files must have been created, e.g., using write_ocr_trainfwrite_ocr_trainfWriteOcrTrainfWriteOcrTrainfWriteOcrTrainfwrite_ocr_trainf, before calling trainf_ocr_class_svmtrainf_ocr_class_svmTrainfOcrClassSvmTrainfOcrClassSvmTrainfOcrClassSvmtrainf_ocr_class_svm. The parameters EpsilonEpsilonEpsilonEpsilonepsilonepsilon and TrainModeTrainModeTrainModeTrainModetrainModetrain_mode have the same meaning as in train_class_svmtrain_class_svmTrainClassSvmTrainClassSvmTrainClassSvmtrain_class_svm。Please, note that training characters that have no corresponding class in the classifier OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle are discarded.

执行信息

多线程类型：可重入（与非独占算子并行运行）。
多线程作用域：全局（可从任何线程调用）。
未采用并行化处理。

此算子修改后续输入参数的状态：

OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle

在执行此算子时，若该参数值需在多个线程间使用，则必须对其访问进行同步。

参数

OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle (输入控制，状态被修改) ocr_svm → (handle)

Handle of the OCR classifier.

TrainingFileTrainingFileTrainingFileTrainingFiletrainingFiletraining_file (输入控制) filename.read(-array) → (string)

Names of the training files.

默认值： 'ocr.trf' "ocr.trf" "ocr.trf" "ocr.trf" "ocr.trf" "ocr.trf"

File extension: .trf, .otr

EpsilonEpsilonEpsilonEpsilonepsilonepsilon (输入控制) real → (real)

Stop parameter for training.

默认值： 0.001

建议值： 0.00001, 0.0001, 0.001, 0.01, 0.1

TrainModeTrainModeTrainModeTrainModetrainModetrain_mode (输入控制) number → (string / integer)

Mode of training.

默认值： 'default' "default" "default" "default" "default" "default"

值列表： 'add_sv_to_train_set'"add_sv_to_train_set""add_sv_to_train_set""add_sv_to_train_set""add_sv_to_train_set""add_sv_to_train_set", 'default'"default""default""default""default""default"

示例（HDevelop）

* Train an OCR classifier
read_ocr_trainf_names ('ocr.trf', CharacterNames, CharacterCount)
create_ocr_class_svm (8, 10, 'constant', 'default', CharacterNames, \
                      'rbf', 0.01, 0.01, 'one-versus-one', \
                      'normalization', 81, OCRHandle)
trainf_ocr_class_svm (OCRHandle, 'ocr.trf', 0.001, 'default')
write_ocr_class_svm (OCRHandle, 'ocr.osc')

结果

如果参数有效，算子 trainf_ocr_class_svmtrainf_ocr_class_svmTrainfOcrClassSvmTrainfOcrClassSvmTrainfOcrClassSvmtrainf_ocr_class_svm 返回值 2 ( H_MSG_TRUE )。如有必要，则抛出异常。

trainf_ocr_class_svmtrainf_ocr_class_svmTrainfOcrClassSvmTrainfOcrClassSvmTrainfOcrClassSvmtrainf_ocr_class_svm may return the error 9211 (Matrix is not positive definite) if PreprocessingPreprocessingPreprocessingPreprocessingpreprocessingpreprocessing = 'canonical_variates'"canonical_variates""canonical_variates""canonical_variates""canonical_variates""canonical_variates" is used. This typically indicates that not enough training samples have been stored for each class. In this case we recommend to change PreprocessingPreprocessingPreprocessingPreprocessingpreprocessingpreprocessing to 'normalization'"normalization""normalization""normalization""normalization""normalization". Another solution can be to add more training samples.

可能的前趋

create_ocr_class_svmcreate_ocr_class_svmCreateOcrClassSvmCreateOcrClassSvmCreateOcrClassSvmcreate_ocr_class_svm, write_ocr_trainfwrite_ocr_trainfWriteOcrTrainfWriteOcrTrainfWriteOcrTrainfwrite_ocr_trainf, append_ocr_trainfappend_ocr_trainfAppendOcrTrainfAppendOcrTrainfAppendOcrTrainfappend_ocr_trainf, write_ocr_trainf_imagewrite_ocr_trainf_imageWriteOcrTrainfImageWriteOcrTrainfImageWriteOcrTrainfImagewrite_ocr_trainf_image

可能的后继

do_ocr_single_class_svmdo_ocr_single_class_svmDoOcrSingleClassSvmDoOcrSingleClassSvmDoOcrSingleClassSvmdo_ocr_single_class_svm, do_ocr_multi_class_svmdo_ocr_multi_class_svmDoOcrMultiClassSvmDoOcrMultiClassSvmDoOcrMultiClassSvmdo_ocr_multi_class_svm, write_ocr_class_svmwrite_ocr_class_svmWriteOcrClassSvmWriteOcrClassSvmWriteOcrClassSvmwrite_ocr_class_svm

替代

read_ocr_class_svmread_ocr_class_svmReadOcrClassSvmReadOcrClassSvmReadOcrClassSvmread_ocr_class_svm

另见

train_class_svmtrain_class_svmTrainClassSvmTrainClassSvmTrainClassSvmtrain_class_svm

模块

光学字符识别/光学字符验证

算子