5

Like, when I see a Chinese character in this picture (link is broken and redirected):

What is the best way to find more information about this character (unlike English, there're only 26 letters, and I can type them and google. Chinese characters cannot be typed if I don't know its pinyin)?

3
  • 1
    In the worst case, you can always use handwriting IME with any dictionary. These input methods are not extremely sensitive to stroke order. Dec 19, 2017 at 4:19
  • this is not an actual Chinese character, instead coined glyph by concatating 脑 and 残。 脑残=foolish. Dec 23, 2017 at 12:38
  • I want it to be translated in English!enter image description here Apr 2, 2021 at 16:29

5 Answers 5

5

With pictures OCR is your best bet.

Here’s a screenshot I took of the front page of baidu just now:

yoyoyo

Then I use Pleco’s OCR software to see what the titles Chinese characters are

looklooklook

here you can see that the characters have been correctly identified.

It works for more than just words too.

Here’s a whole headline deciphered:

yup

But in Pleco they can even look up the words one by one for you:

fancy right?

Fancy! Right?

——

Method number two: Handwriting Recognition

...and again we’re using Pleco (and really why would you need any other software? Really?)

So let’s stick with out original picture - we don’t know the characters for baidu.

So first we write our bai.

hahaha

Then we write our du:

yes

and now we have our two characters! And thankfully we’re using dictionary software so we have plenty of definitions:

un-hunh


Method three: two part character search

We're getting more advanced now.

So let's take the 娱 from 娱乐圈 in the OCR example above.

Now we know a little bit about Chinese characters - we don't know 娱 yet - but we know it's comprised of 女 and 吴.

So what can we do?

This is were zisea comes in handy. We can use their 两分 (or "two part") character system to find what we're looking for:

yuss

So in the "two part" search box we have:

女吴

then we see our results:

how?

and there we see our 娱 right at the very top with a brief description!

if we click into it we get a page like this:

yo!

with the character, pinyin, descriptions, definitions and alternative characters.


Method number four: 笔顺

This is, really, not my favorite - but I've used it enough times to make to worth mentioning.

If OCR, handwriting recognition and two-part searches are failing you - you can try bishun.

The rules for bishun are simple:

oooo

Solely based off how to write the character we can convert it to a number that can be searched.

So let's take: 乐 we don't know to read it but it's stroke order is obvious:

ノフ丨ノ丶

so according to the above table that gives us:

35234

Now if we go to zdic.net

and click on the second tab:

字典

like this:

like this

then click on:

笔顺

like so:

noa

now we can enter out number

![enter image description here

and here are our results:

uop

with:

搜索结果:

乐 35234

right at the top


All of these methods would work with any character you don't know.

As long as you have a clear enough picture you can run OCR on any character.

While handwriting, two-part search and bishun methods can also work with any characters you find out there in the wild.

6
  • Amazing! I wish I could do that!
    – Pedroski
    Dec 17, 2017 at 23:13
  • Thanks! But I still cannot correctly recognize the character in the picture placed in this question. Is that really an existing character?
    – ice1000
    Dec 18, 2017 at 4:43
  • na, that's why @BirchBryant said, "Well, the first step is for the character in the picture to actually be 汉字。"
    – Mou某
    Dec 18, 2017 at 4:44
  • "really why would you need any other software?" well, many people say that Hanping Camera (disclaimer: I'm the developer) is by far the best Chinese OCR app on Android: chinese-forums.com/forums/topic/55434-pleco-ocr-attn-mike-love/… and regarding taking screenshots - no need when you have Hanping Popup! youtube.com/watch?v=ox0p0YgeJkI
    – Mark
    Dec 19, 2017 at 4:02
  • @user3306356 When you see text "on a picture" OCR is most of the time useless because of the inconsistent background. IMO the answer would be much better (more compact) if you removed the OCR section.
    – Mark
    Dec 22, 2017 at 13:28
3

The word in the picture is a combination of two words, 脑 and 残, as shown in the image. Somebody made it up.

enter image description here

脑残 is a new phrase in the modern world, so I don't think you can find any method to input the word in the picture.

If you know the Cangjie input method (倉頡輸入法), the input keys of 脑 are 月b, 卜y, 山u and 大k. And, those of 残 are 一m, 弓n, 戈i and 手q.

According to the rules of the Cangjie IM, the input keys of the word in the picture are 一m, 月b, 戈i, 十j and 山u.

0
2

In this case, the background is too noisy/complex for it be even worth giving OCR a try. Instead, use any Chinese IME (input method editor/soft keyboard) that supports handwriting recognition and hope that it is forgiving enough to allow for errors in stroke order.

If the background is clean and the font not too stylised, then a Chinese OCR would probably be the best option.

2

Another way is to interpret the first two strokes and the last two strokes as numbers. so bai 百 is 1600 and du 度 is 3676. This system was first proposed in the shou wei hao ma 首尾號碼字典 dictionary by 陳舜齊 in Taiwan.

1

I would use Google Keep. You take a picture on the street or cut an image from the screen. Paste into a new Keep note on your computer. Then you can "Grab image text". Then you can translate it into English on Google Translation.

For example, below is my Keep note. I cut this image from one of the answers to this question, and I get the Chinese and English text.

enter image description here

(*I not not sure how it works on Phones)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.