惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

WordPress大学
WordPress大学
大猫的无限游戏
大猫的无限游戏
酷 壳 – CoolShell
酷 壳 – CoolShell
博客园 - 叶小钗
月光博客
月光博客
Last Week in AI
Last Week in AI
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
人人都是产品经理
人人都是产品经理
阮一峰的网络日志
阮一峰的网络日志
罗磊的独立博客
IT之家
IT之家
美团技术团队
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
Hugging Face - Blog
Hugging Face - Blog
博客园_首页
S
SegmentFault 最新的问题
宝玉的分享
宝玉的分享
博客园 - Franky
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
Jina AI
Jina AI
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
The Cloudflare Blog
博客园 - 司徒正美
爱范儿
爱范儿

LWN.net comments

tcmalloc's weird hack [LWN.net] Fixed? [LWN.net] mpd [LWN.net] Userspace AX.25 [LWN.net] RIP [LWN.net] My two cents... [LWN.net] pipx [LWN.net] Tragedy [LWN.net] A young man destined for glory [LWN.net] And 'less' won't let you search [LWN.net] A great loss [LWN.net] Sad and shocking news [LWN.net] Easy migration from Clementine [LWN.net] Sad coincidence [LWN.net] GNOME is actually usable thanks to Seth et al [LWN.net] Sad news :( [LWN.net] armhf supports preempt_rt [LWN.net] MusicBrainz accurracy [LWN.net] On open source maintainership [LWN.net] Let's stop here [LWN.net] Not a new thing [LWN.net] uv is indeed great pgmoneta Some comments on this on a Postgres blog feed [LWN.net] uv [LWN.net] going to Debian [LWN.net] Upgrading 64-bit-capable systems to 64-bit kernels? [LWN.net] Free Software foundations Maintainers can wait for code review but not for publish review? A reasonably extreme point of view [LWN.net]
Tesseract does need 'clean-ish' documents [LWN.net]
fest3er · 2026-06-18 · via LWN.net comments

Tesseract does need 'clean-ish' documents

Posted Jun 18, 2026 13:53 UTC (Thu) by fest3er (guest, #60379)
In reply to: Tesseract does need 'clean-ish' documents by vasvir
Parent article: FairScan 2.0 released

Sorry, can't help with CamScanner; this is the first I've heard of it.

In my case, most of the scans were similar. Once I figured out the necessary adjustments needed by using Gimp, I then used ImageMagick to process all of the scans (mostly adjusting gamma. I also wrote a script that let me use all available CPUs to process that many images at a time using ImageMagick, and another using Tesseract to OCR all those images using all CPUs. In short, I used Gimp to figure out what needed to be done to all of them, and ImageMagick to apply those adjustments to all of the scans.