Malcon的OCR字符识别应用起来很方便，可以保存和读取，
1、读取图片ReadImage
2、阈值分割Threshold
3、连通区域Connection
4、排序区域SortRegion
5、贴上类标签Class:=[\'a\',\'a\',\'a\',\'a\',\'b\',\'b\',\'b\',\'b\',\'c\',\'c\',\'c\',\'c\']
6、写入字符文件WriteOcrTrainf
7、读取字符文件：输出标签类名和数量ReadOcrTrainfNames
8、创建OCR句柄CreateOcrClassKnn
9、训练OCR句柄TrainfOcrClassKnn
10、单个字符分类DoOcrSingleClassKnn
11、多个字符分类DoOcrMultiClassKnn
12、清除字符句柄ClearOcrClassKnn
13、获取字符区域的位置坐标AreaCenter
14、设置显示字体SetDisplayFont
15、显示字符信息DispMessage
ReadImage(moImage, \'a1.bmp\')
Threshold(moImage, thRegion, 0, 50)
Connection(thRegion, thRegion)
//区域从左到右，从上到下排序
SortRegion(thRegion, StRegion,\'character\', \'true\', \'row\')
Class:=[\'a\',\'a\',\'a\',\'a\',\'b\',\'b\',\'b\',\'b\',\'c\',\'c\',\'c\',\'c\']
WriteOcrTrainf(StRegion,moImage,Class,\'KNNletters.trf\')
ReadOcrTrainfNames(\'KNNletters.trf\',name,count)
CreateOcrClassKnn(15,15,\'\',\'\',name,\'\',\'\',ocrHandle)
TrainfOcrClassKnn(ocrHandle,\'KNNletters.trf\',\'\',\'\')
//开始分类
DoOcrSingleClassKnn(StRegion,moImage,ocrHandle,1,1,Class2,Confidence)
DoOcrMultiClassKnn(StRegion,moImage,ocrHandle,Class3,Confidence2)
ClearOcrClassKnn(ocrHandle)
//把结果显示出来
AreaCenter(StRegion,Area,Row,Column)
SetDisplayFont(3600,40,\'\',\'true\',\'false\')
DispMessage(3600,Class3,\'\',Row,Column,\'red\',\'false\')