ttop32/jmtrans

manga translator - get japanese manga from url to translate manga image

jmtrans's Issues

Is this free I do not want to pay for anything :D

It uses google cloud i guess

Program alters the original image when doing text segmentation

When using SickZil to find supposed text on the image, the supposed text is erased and the image is altered to match the background.
Then, the supposed text is processed with OpenCV to detect the real text and translate it.

However, the supposed text detected by SickZil that isn't identified as text by OpenCV is erased from the original image. Most of the time this means that some strokes, random strands of hair or small text are deleted from the original image.

But other times entire objects or facial features are deleted from the image, such as here, where the right eye of the character is missing in the translated version, as well as some details on her hair:

A solution could be changing the aggressiveness of SickZil if that's possible since it seems it's too high right now.

Or you could just reintegrate those parts of the image picked up by SickZil but not recognised as text by OpenCV. That way, in the case of the images I've uploaded, the eye and hair details would be extracted from the image but then reintegrated when OpenCV doesn't recognise them.

In detect text and translate, the iterations time increases rapidly .

 19%|█▉        | 9/47 [12:19<1:05:20, 103.18s/it]

What is the possible reason for it and any way to avoid it .

Json error

The prpblem is in this part of code :

translator.translate(text, lang_tgt=langCode)   if len(text)!=0 else "" 

Error :

JSONDecodeError                           Traceback (most recent call last)
<ipython-input-25-15d7d09b1ca0> in <module>()
     17     textList_trans=[]
     18     for text in textList:
---> 19         text_trans=translator.translate(text, lang_tgt=langCode,)    if len(text)!=0 else ""
     20         textList_trans+=[text_trans]
     21     textListDict_trans[fileName]=textList_trans

3 frames
/usr/lib/python3.7/json/decoder.py in decode(self, s, _w)
    338         end = _w(s, end).end()
    339         if end != len(s):
--> 340             raise JSONDecodeError("Extra data", s, end)
    341         return obj
    342 

JSONDecodeError: Extra data: line 1 column 352 (char 351)

eztrans로 번역을 선택하고 실행하면 오류가 일어납니다

download done
progess
0%| | 0/10 [00:00<?, ?it/s]Exception in thread Thread-5:
Traceback (most recent call last):
File "threading.py", line 914, in _bootstrap_inner
File "threading.py", line 862, in run
File "manga_translator.py", line 97, in processTranslationTask
File "text_translate.py", line 39, in translate
File "text_translate.py", line 16, in translateList
File "text_translate.py", line 26, in translateEztrans
File "subprocess.py", line 676, in init
File "subprocess.py", line 957, in _execute_child
FileNotFoundError: [WinError 2] 지정된 파일을 찾을 수 없습니다

google ocr 과 eztrans 를 사용하였을때 이런 오류가 계속 일어납니다

Supported site url is dead

"Download failed. May be Goodbye DPI not on."

Sorry for the simple issue. It showed up when I put the site url inside the link bar. Wasn't sure is it because of the invalid site link or the other problem. The picture is the ss of my GoodbyeDPI.

Note: the supported site url list is dead.

English translation text overlaps sometimes

When translating from Japanese to English, on some occasions the translated text overlaps.

Here's an example where the translation has overlaping text on the left and correctly formated text on the right:

I think that the problem may occur whenever the original text is written vertically (in languages such as with Chinese, Japanese or Korean) and the destination language only supports horizontal writing (such as English, Spanish or French).

Checking the size of the new text boxes, testing if they will overlap and displacing the text downwards could be a solution, since I think it's preferable having the text protrude from the bottom of the speech bubble than having it overlap.