-
# 预览截图:
![image](https://github.com/hiroi-sora/Umi-OCR/assets/56373419/fc2266ee-b9b7-4079-8b10-6610e6da6cf5)
## 扫描件转为双层可搜索PDF 示例:
![image](https://github.com/hiroi-sora/Umi-OCR/assets/5637341…
-
建议以后打包发行时不要用7zip
7zip的压缩包不会保存unix系统的文件权限,这很容易出错。
![image](https://github.com/hiroi-sora/PaddleOCR-json/assets/71205842/552dcbc5-69af-480c-a082-9bde852ad1e4)
像上面这样,7zip发行压缩包在解压后丢失了PaddleOCR-js…
-
用这张图测试,旮旯二字只识出一个:
-
![Snipaste_2023-11-08_23-39-24](https://github.com/hiroi-sora/Umi-OCR_v2/assets/69130137/b18d7d7c-c47b-477c-8b79-4d13f858feeb)
如上图,接口和软件中的不一样,我理解返回的数据应该是要处理,但是并不清楚具体的逻辑
-
https://github.com/hiroi-sora/Umi-OCR/issues/160
https://private-user-images.githubusercontent.com/56373419/246670568-8b4da183-d4f9-4f06-b39f-2fab85d37d80.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9…
-
识别的英文单词之间没有空格了怎么解决?
-
![image](https://github.com/hiroi-sora/Umi-OCR/assets/22903427/46aba856-7a46-4e9b-9739-cfab5ebe97e4)
背景是批量的识别图片,用CLI,出现了这个错误
-
建议截图识别也可以保存结果到本地指定的文件
cukkk updated
5 months ago
-
-
![截图_20240313211355](https://github.com/hiroi-sora/Umi-OCR/assets/107912895/3eca23dc-03bc-46ba-814a-3866b639f4f3)
生成后是0KB打不开,什么原因呢