xiaoyifang / goldendict-ng

The Next Generation GoldenDict
https://xiaoyifang.github.io/goldendict-ng/
Other
1.7k stars 95 forks source link

Is it possible to pick up the text on the picture? #1248

Closed buddhiko1 closed 1 year ago

buddhiko1 commented 1 year ago

首先,再次感谢这个开源项目的所有贡献人员,GoldenDict是我日常生活和工作中不可或缺的一个工具。

最近我在pc上读一些pdf格式的古籍(如下图),经常需要查字典。因为文字是在图片上的,所以没办法选择文字。每次都要拿出手机,然后在欧陆上手写查询,非常的耗时耗力。

对于图片上的文字,能否通过hover去拾取呢?是否有实现的可能性?因为有很多pdf都是图片形式的,这个问题我想很多人都会遇到。

last

github-actions[bot] commented 1 year ago

Bot detected the issue body's language is not English, translate it automatically.

First of all, I would like to thank all the contributors to this open source project again. GoldenDict is an indispensable tool in my daily life and work.

Recently, I am reading some ancient books in pdf format on my PC (as shown below), and I often need to look them up in a dictionary. Because the text is on the image, there is no way to select the text. Every time I have to take out my mobile phone and handwrite the query on the European continent, it is very time-consuming and labor-intensive.

Can the text on the picture be picked up through hover? Is it possible to achieve this? Because many PDFs are in the form of pictures, I think many people will encounter this problem.

last

buddhiko1 commented 1 year ago

应该没有可能识别这种古籍字体。

github-actions[bot] commented 1 year ago

Bot detected the issue body's language is not English, translate it automatically.

It should not be possible to recognize this estimated font.

xiaoyifang commented 1 year ago

https://xiaoyifang.github.io/goldendict-ng/howto/ocr/

buddhiko1 commented 1 year ago

https://xiaoyifang.github.io/goldendict-ng/howto/ocr/

非常感谢! 帮了我的大忙. 给capture2text安装额外的繁体语言包后亲测可用,识别率还挺高.

github-actions[bot] commented 1 year ago

Bot detected the issue body's language is not English, translate it automatically.

https://xiaoyifang.github.io/goldendict-ng/howto/ocr/

Thank you very much! It helped me a lot. After installing an additional Traditional Chinese language package for capture2text, I personally tested it and found that the recognition rate is quite high.