为Tesseract OCR创建训练图像 [英] Creating a training image for Tesseract OCR

问题描述

我正在编写用于为Tesseract OCR训练图像的生成器.

I'm writing a generator for training images for Tesseract OCR.

为Tesseract OCR的新字体生成训练图像时，最佳值是:

When generating a training image for a new font for Tesseract OCR, what are the best values for:

1. DPI
2. 以磅为单位的字体大小
3. 字体是否应消除锯齿
4. 边界框是否应该紧贴:，否则:

推荐答案

我找到了第四个问题的答案-边界框是否应该紧密贴合".

I found the answer to the 4th question - "Should the bounding boxes fit snugly".

似乎尽可能地拟合矩形会带来更好的结果.

It seems that fitting the rectangles as much as possible gives much better results.

对于其他12个点和300 dpi，就足够了，如@Yaroslav所建议的那样.我认为最好关闭抗锯齿功能.

For the other 12 pts and 300 dpi will be good enough, as @Yaroslav suggested. I think anti-aliasing is better turned off.

这篇关于为Tesseract OCR创建训练图像的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！