c# - 如何让Tesseract OCR输出句子形式的单词？

Question

我得到的结果是这样的: http://i.stack.imgur.com/dM0qG.png

是否可以让 Tesseract 以这样的句子/段落形式输出？

This is to certify that you have successfully PASSED the PHIL-IT General Certification Examination held on January 26, 2015 at the Cebu Institute of Technology - University, N. Bacalso Avenue, Cebu City 6000 Philippines.

Answer 1

因为result是一个Tessnet2.Word的List，并且存储了每个Word的文本在其 item.Text 中，您可以:

1. 创建一个只有单词的列表(不是完整的 Tessnet2.Word 对象)
2. 加入这个列表，使用“空格”作为分隔符

假设您的结果存储在名为 result 的 var 中(您执行了操作
var result = ocr.DoOCR(image, null);).如果结合这两个步骤，它看起来像这样:

string phrase = string.Join(" ", result.Select(x => x.Text).ToList());

结果是:

This is to certify that you have successfully PASSED the Phil-lT General Certification Examination held on [nnuag 26, 2015 at the Cebu Institute uf Tedmnlngy · University , N. Bacalso Avenue, Cebu City 6000 Philippines.

(它有一些检测错误，但那是另一个问题)