# Scan in some \[hand\]written Traditional Chinese **URL:** \ **Category:** Technology\ **Created:** [January 25, 2023, 6:09am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935 "2023-01-25T06:09:10Z")\ **Posts on this page:** 17\ **Page:** 1
**Author:** ![bdog](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/b/919ad9/32.png) [@bdog](https://tw.forumosa.com/u/bdog)\ **Post date:** [January 25, 2023, 6:09am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/1 "2023-01-25T06:09:10Z")
How to take a photo of some written Chinese and have it show up on the screen without, preferably, using any A Pee Pees? Recommendations for the best web based OCR tool basically. Thanks. ---
**Author:** ![Marco](https://forumosa-12829.kxcdn.com/user_avatar/tw.forumosa.com/marco/32/179744_2.png) [@Marco](https://tw.forumosa.com/u/Marco)\ **Post date:** [January 25, 2023, 7:24am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/2 "2023-01-25T07:24:35Z")
Google Translate. Chinese to Chinese Traditional ---
**Author:** ![Charlie\_Jack](https://forumosa-12829.kxcdn.com/user_avatar/tw.forumosa.com/charlie_jack/32/246096_2.png) [@Charlie\_Jack](https://tw.forumosa.com/u/Charlie_Jack)\ **Post date:** [January 25, 2023, 5:23pm UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/3 "2023-01-25T17:23:14Z")
> [@bdog](#): > > How to take a photo of some written Chinese and have it show up on the screen without, preferably, using any A Pee Pees? Recommendations for the best web based OCR tool basically. Thanks. Caveat: I make no claims; in fact I’m not even sure I know what you’re talking about. In the past, I tried an online OCR device for an image (or images) of English text (or a device that was downloadable–it was so long ago that I don’t remember exactly what kind of device it was), and the result was an unintentionally-humorous failure ([ah, here’s a post with a quote of the failure](https://tw.forumosa.com/t/vocab-list-fer-passin-the-high-school-entrance-exam/44695/54)). But after an ordinary Google search, I just now found [this device](https://www.i2ocr.com/features). I tested it with an image of a few Traditional Chinese characters screen-captured from a Google Translate page using IrfanView image software. (The text is a Chinese translation of a small part of Abraham Lincoln’s Gettysburg Address–I didn’t use Google Translate for the translation–it had already been translated.) For **those few** characters, it **seems** to have worked **that one time**. That’s all I can say. Here’s an image of the results: ![image](https://forumosauploads-12829.kxcdn.com/original/3X/7/f/7f89152d292f93a77cc02882646b8c2bed3a33c0.jpeg) But again, I make no guarantees of any kind about this device, and that also means that I can’t guarantee that there’s nothing harmful about the device. Anyway, I hope this helps, or at least does no harm. ---
**Author:** ![nz](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/n/a9adbd/32.png) [@nz](https://tw.forumosa.com/u/nz)\ **Post date:** [January 25, 2023, 11:55pm UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/4 "2023-01-25T23:55:52Z")
Pleco’s “still image” OCR function is pretty good, assuming you have good lighting and you’re not trying to scan a book that has text too close to the center (warped image). It still gets similar characters wrong sometimes (估 vs 佔 was driving me nuts last week), but otherwise it gets most of the text right. ---
**Author:** ![bdog](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/b/919ad9/32.png) [@bdog](https://tw.forumosa.com/u/bdog)\ **Post date:** [January 26, 2023, 2:27am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/5 "2023-01-26T02:27:47Z")
> [@Charlie\_Jack](#): > > or at least does no harm It’s alright I’m not translating nuclear launch codes. Well not yet anyway, who knows what the future holds for me here. ---
**Author:** ![bdog](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/b/919ad9/32.png) [@bdog](https://tw.forumosa.com/u/bdog)\ **Post date:** [January 26, 2023, 2:30am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/6 "2023-01-26T02:30:37Z")
Thanks, you appear to love Pleco. Unfortunately my phone has what I am coining as ‘Android cancer’. The disk usage just keeps growing, I keep deleting apps, now there is not enough space to upgrade or do anything. Pleco is the last and only app I have left, I refuse to delete it because it is so useful. Perhaps it is the problem? ---
**Author:** ![nz](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/n/a9adbd/32.png) [@nz](https://tw.forumosa.com/u/nz)\ **Post date:** [January 26, 2023, 11:54am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/7 "2023-01-26T11:54:19Z")
I hope I’m not the only person on here who loves Pleco. If you’re going to need one app for Chinese survival, Pleco would be the one. But I would imagine it takes up an enormous amount of space — it’s all offline, including the OCR. Maybe contact the Pleco people and ask? I know they’ve posted on [Chinese Forums](https://www.chinese-forums.com/), but you might also get to them faster with an email ---
**Author:** ![bdog](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/b/919ad9/32.png) [@bdog](https://tw.forumosa.com/u/bdog)\ **Post date:** [January 27, 2023, 2:22am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/8 "2023-01-27T02:22:47Z")
Anyone with some further recommendations for OCR on handwritten input? Many thanks. ---
**Author:** ![Charlie\_Jack](https://forumosa-12829.kxcdn.com/user_avatar/tw.forumosa.com/charlie_jack/32/246096_2.png) [@Charlie\_Jack](https://tw.forumosa.com/u/Charlie_Jack)\ **Post date:** [January 27, 2023, 6:41am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/9 "2023-01-27T06:41:14Z")
> [@bdog](#): > > Anyone with some further recommendations for OCR on handwritten input? Many thanks. I _think_ that _maybe_ Google Lens can _sorta_ do that _sometimes_. 🙂 Or maybe it can do it regularly, but I don’t know. ![image](https://forumosauploads-12829.kxcdn.com/original/3X/4/7/47978bc5654a0efce19f7a601ebb1b00e9be052b.jpeg) For Chinese, that was a first-time thing for me. Earlier I tried handwritten English (a soldier writing a letter home), and it worked pretty well. Then I realized I’d read too hastily and misunderstood your post, so I went looking for handwritten Chinese. Some time ago I tried it on something that looked like Arabic, but the image was pretty indistinct, if I remember correctly. But I’m guessing you want something that can do a lot of it at one time. I’m not sure about that. This is terra incognita for me. I wish I could be of better assistance, but that’s all I’ve got. Hopefully, someone will come up with something better. ---
**Author:** ![bdog](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/b/919ad9/32.png) [@bdog](https://tw.forumosa.com/u/bdog)\ **Post date:** [January 27, 2023, 6:44am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/10 "2023-01-27T06:44:34Z")
Thanks, apologies I was unclear, reading it back. I’m actually surprised recognition of handwriting isn’t further along, technologically I mean. ---
**Author:** ![Charlie\_Jack](https://forumosa-12829.kxcdn.com/user_avatar/tw.forumosa.com/charlie_jack/32/246096_2.png) [@Charlie\_Jack](https://tw.forumosa.com/u/Charlie_Jack)\ **Post date:** [January 27, 2023, 6:46am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/11 "2023-01-27T06:46:28Z")
> [@bdog](#): > > Thanks, apologies I was unclear, reading it back. I’m actually surprised recognition of handwriting isn’t further along, technologically I mean. No apology necessary. Really I should apologize. I’m really afraid I’m wasting your time, because you have practical concerns. ---
**Author:** ![TW1](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/t/f4b2a3/32.png) [@TW1](https://tw.forumosa.com/u/TW1)\ **Post date:** [January 27, 2023, 2:46pm UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/12 "2023-01-27T14:46:36Z")
Not exactly what is requested here, but I found this OCR tool recently which helped me a lot (on MacOS). > **[iText - OCR Tool](https://en.toolinbox.net/iText/)** > > iText is an OCR tool which could recognize text from any image. You can use iText to extract text from PDF, document in paper, page in a book and any other images. 1. Easily Select ImageiText support You can “make a screenshot” of a section of your screen, and any character in the screenshot will be copied to the clipboard and translated in a menubar popup. This is different from Google Translate, because this tool allows you to recognize the characters from a sentence (not selectable) inside a picture for example. (Similar to Lens, but on your computer instead of using your phone camera) There are many tools and add-ons that allow to translate (with a popup, etc.) text that can be selected on a webpage, but this is the first one I found that works well for the text inside an image. You have to pay however, about $10-$15/Y to get unlimited translations, otherwise you get a limited amount. After using it for a few days, I bought the yearly sub. Very helpful for me. (Shopee ads with Chinese text embedded in pictures for example) ---
**Author:** ![teosocrates](https://forumosa-12829.kxcdn.com/user_avatar/tw.forumosa.com/teosocrates/32/264158_2.png) [@teosocrates](https://tw.forumosa.com/u/teosocrates)\ **Post date:** [January 2, 2025, 10:55pm UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/13 "2025-01-02T22:55:05Z")
Most newer iphones can screenshot and convert text now; chatgpt also does a pretty great job of picking up handwritten notes and transcribing them. ---
**Author:** ![shaokilai](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/s/eada6e/32.png) [@shaokilai](https://tw.forumosa.com/u/shaokilai)\ **Post date:** [January 2, 2025, 11:15pm UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/14 "2025-01-02T23:15:32Z")
ChatGPT - upload a jpg image and ask it to transcribe, then if you need it translate to your language of choice ---
**Author:** ![bz](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/b/ec9cab/32.png) [@bz](https://tw.forumosa.com/u/bz)\ **Post date:** [January 3, 2025, 3:16am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/15 "2025-01-03T03:16:49Z")
> [@shaokilai](#): > > ChatGPT It’s brand new, but (so far) Google’s Gemini Flash 2 beats all the OpenAI options handily on the OCR front. 🤷 ---
**Author:** ![teosocrates](https://forumosa-12829.kxcdn.com/user_avatar/tw.forumosa.com/teosocrates/32/264158_2.png) [@teosocrates](https://tw.forumosa.com/u/teosocrates)\ **Post date:** [January 17, 2026, 9:02am UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/16 "2026-01-17T09:02:24Z")
I actually bought those Rokid glasses… I hoped it would translate and read things out (my intermediate books don’t have pinyin so it’s slow to memorize every character first). The glasses can take a screenshot and translate, but not also read them; with chatgpt you can take a picture, then ask chatgpt in voice mode to read it or practice discussing the material. That stuff works but it was also too slow and manual, so I built a whole platform just with my study materials and lessons, vocab flashcards etc - I’m hoping it’ll finally be what I need to improve. ---
**Author:** ![jojolingo](https://forumosa-12829.kxcdn.com/letter_avatar_proxy/v4/letter/j/839c29/32.png) [@jojolingo](https://tw.forumosa.com/u/jojolingo)\ **Post date:** [October 6, 2026, 4:17pm UTC](https://tw.forumosa.com/t/scan-in-some-hand-written-traditional-chinese/226935/17 "2026-10-06T16:17:40Z")
@teosocrates Totally feel you on the manual capture process being too slow to stay in the flow. Pausing every few seconds to take screenshots or draw 20-stroke Traditional characters into Pleco kills video immersion completely. I hit that exact wall with Taiwanese YouTube channels and dramas with burned-in subs, so I built a browser extension called **YuzuLingo**. It runs client-side OCR directly over the web video player so mouseover dictionaries (like Zhongwen or Yomitan) work on hover over hardsubs in real-time. Zero API keys or local Python setups required. Works best on standard broadcast fonts (variety show graphics are still hit-or-miss), but if you watch a lot of web video for Traditional Chinese practice, it might save you some time.