do_ocr_single_class_cnnT_do_ocr_single_class_cnnDoOcrSingleClassCnnDoOcrSingleClassCnndo_ocr_single_class_cnn (算子)

名称

do_ocr_single_class_cnnT_do_ocr_single_class_cnnDoOcrSingleClassCnnDoOcrSingleClassCnndo_ocr_single_class_cnn — 使用基于 CNN 的 OCR 分类器对单个字符进行分类。

签名

do_ocr_single_class_cnn(Character, Image : : OCRHandle, Num : Class, Confidence)

描述

do_ocr_single_class_cnndo_ocr_single_class_cnnDoOcrSingleClassCnnDoOcrSingleClassCnnDoOcrSingleClassCnndo_ocr_single_class_cnn computes the best classification for the character given by the region CharacterCharacterCharacterCharactercharactercharacter and the gray values ImageImageImageImageimageimage with the OCR classifier OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle and returns the best NumNumNumNumnumnum classes in ClassClassClassClassclassValclass and the corresponding confidences (probabilities) of the classes in ConfidenceConfidenceConfidenceConfidenceconfidenceconfidence. Because multiple classes may be returned by do_ocr_single_class_cnndo_ocr_single_class_cnnDoOcrSingleClassCnnDoOcrSingleClassCnnDoOcrSingleClassCnndo_ocr_single_class_cnn, CharacterCharacterCharacterCharactercharactercharacter may only contain a single region (a single character). If multiple characters should be classified in a single call, do_ocr_multi_class_cnndo_ocr_multi_class_cnnDoOcrMultiClassCnnDoOcrMultiClassCnnDoOcrMultiClassCnndo_ocr_multi_class_cnn must be used.

In most cases, do_ocr_multi_class_cnndo_ocr_multi_class_cnnDoOcrMultiClassCnnDoOcrMultiClassCnnDoOcrMultiClassCnndo_ocr_multi_class_cnn should be preferred over do_ocr_single_class_cnndo_ocr_single_class_cnnDoOcrSingleClassCnnDoOcrSingleClassCnnDoOcrSingleClassCnndo_ocr_single_class_cnn, since do_ocr_multi_class_cnndo_ocr_multi_class_cnnDoOcrMultiClassCnnDoOcrMultiClassCnnDoOcrMultiClassCnndo_ocr_multi_class_cnn is typically faster than a loop with do_ocr_single_class_cnndo_ocr_single_class_cnnDoOcrSingleClassCnnDoOcrSingleClassCnnDoOcrSingleClassCnndo_ocr_single_class_cnn. Furthermore, the resulting confidences can be interpreted as probabilities. They indicate the uncertainty of a character's classification result. The operator do_ocr_single_class_cnndo_ocr_single_class_cnnDoOcrSingleClassCnnDoOcrSingleClassCnnDoOcrSingleClassCnndo_ocr_single_class_cnn has to be used if the nth-best class has to be examined explicitly for n greater than one.

A string of the number '\032'"\032""\032""\032""\032""\032" (alternatively displayed as '\0x1A'"\0x1A""\0x1A""\0x1A""\0x1A""\0x1A") in ClassClassClassClassclassValclass signifies that the region has been classified as rejection class.

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。

参数

CharacterCharacterCharacterCharactercharactercharacter (输入对象) region → object

Character to be recognized.

ImageImageImageImageimageimage (输入对象) singlechannelimage → object (byte / uint2)

Gray values of the character.

OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle (输入控制) ocr_cnn → (handle)

Handle of the OCR classifier.

NumNumNumNumnumnum (输入控制) integer-array → (integer)

Number of best classes to determine.

默认值： 1

建议值： 1, 2, 3, 4, 5

ClassClassClassClassclassValclass (输出控制) string(-array) → (string)

Result of classifying the character with the CNN.

ConfidenceConfidenceConfidenceConfidenceconfidenceconfidence (输出控制) real(-array) → (real)

Confidence(s) of the class(es) of the character.

示例（HDevelop）

read_image(Image, 'bottle2')
OffsetRow := 100
OffsetCol := 108
read_ocr_class_cnn('Universal_0-9_NoRej.occ', OCRHandle)
* Select each digit and use do_ocr_single_class_cnn to apply OCR
gen_rectangle1(ROI_Date, OffsetRow, OffsetCol, OffsetRow, OffsetCol)
for I := 1 to 6 by 1
  Offset := I % 2 * 10
  smallest_rectangle1(ROI_Date, Row1, Col1, Row2, Col2)
  gen_rectangle1(ROI_Date, OffsetRow, Offset + Col2, OffsetRow + 42, \
  Col2 + Offset + 31)
  do_ocr_single_class_cnn(ROI_Date, Image, OCRHandle, 1, Class, Confidence)
endfor

结果

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

可能的前趋

read_ocr_class_cnnread_ocr_class_cnnReadOcrClassCnnReadOcrClassCnnReadOcrClassCnnread_ocr_class_cnn

替代

do_ocr_multi_class_cnndo_ocr_multi_class_cnnDoOcrMultiClassCnnDoOcrMultiClassCnnDoOcrMultiClassCnndo_ocr_multi_class_cnn

模块

光学字符识别/光学字符验证

算子