manga image translator
beta-0.3
翻译漫画/图像中的文本。
中文说明 |变更日志
加入我们的不和谐 https://discord.gg/Ak8APNy4vb
有些漫画/图像永远不会被翻译，因此这个项目诞生了。
请注意，示例可能并不总是更新，它们可能不代表当前的主分支版本。
| 原来的 | 已翻译 |
|---|---|
 （来源@09ra_19ra） |  （面具） |
 （来源@VERTIGRIS_ART） |  --detector ctd （掩模） |
 （来源@hiduki_yayoi） |  --translator none （掩码） |
 （来源@rikak） |  （面具） |
官方演示（by zyddnys）：https://touhou.ai/imgtrans/
浏览器用户脚本（由 QiroNT 提供）：https://greasyfork.org/scripts/437569
MMDOCR-高性能的后继者。
这是一个业余爱好项目，欢迎您贡献！
目前这只是一个简单的demo，还存在很多不完善的地方，我们需要您的支持来让这个项目变得更好！
主要设计用于翻译日语文本，但也支持中文、英语和韩语。
支持修复、文本渲染和着色。
# First, you need to have Python(>=3.8) installed on your system
# The latest version often does not work with some pytorch libraries yet
$ python --version
Python 3.10.6
# Clone this repo
$ git clone https://github.com/zyddnys/manga-image-translator.git
# Create venv
$ python -m venv venv
# Activate venv
$ source venv/bin/activate
# For --use-gpu option go to https://pytorch.org/ and follow
# pytorch installation instructions. Add `--upgrade --force-reinstall`
# to the pip command to overwrite the currently installed pytorch version.
# Install the dependencies
$ pip install -r requirements.txt模型将在运行时下载到./models中。
在开始 pip 安装之前，首先安装 Microsoft C++ 构建工具（下载、说明），因为如果没有它，某些 pip 依赖项将无法编译。 （参见#114）。
要在 Windows 上使用 cuda，请按照 https://pytorch.org/ 上的说明安装正确的 pytorch 版本。
要求：
demo/doc文件夹中的文件，则可选）该项目在zyddnys/manga-image-translator:main image 下有 docker 支持。此 docker 映像包含项目所需的所有依赖项/模型。应该注意的是，该图像相当大（~ 15GB）。
可以使用（对于 CPU）托管 Web 服务器
docker run -p 5003:5003 -v result:/app/result --ipc=host --rm zyddnys/manga-image-translator:main -l ENG --manga2eng -v --mode web --host=0.0.0.0 --port=5003或者
docker-compose -f demo/doc/docker-compose-web-with-cpu.yml up取决于你更喜欢哪一个。 Web 服务器应在端口 5003 上启动，并且图像应位于/result文件夹中。
将 docker 与 CLI 结合使用（即在批处理模式下）
docker run -v < targetFolder > :/app/ < targetFolder > -v < targetFolder > -translated:/app/ < targetFolder > -translated --ipc=host --rm zyddnys/manga-image-translator:main --mode=batch -i=/app/ < targetFolder > < cli flags >注意：如果您需要引用主机上的文件，则需要将关联的文件作为卷安装到容器内的/app文件夹中。 CLI 的路径需要是内部 docker 路径/app/...而不是主机上的路径
某些翻译服务需要 API 密钥才能发挥作用，以将它们作为环境变量传递到 docker 容器中。例如：
docker run --env= " DEEPL_AUTH_KEY=xxx " --ipc=host --rm zyddnys/manga-image-translator:main < cli flags > 要与受支持的 GPU 一起使用，请首先阅读初始
Docker部分。您需要使用一些特殊的依赖项
要运行设置了以下标志的容器：
docker run ... --gpus=all ... zyddnys/manga-image-translator:main ... --use-gpu或者（对于 Web 服务器 + GPU）
docker-compose -f demo/doc/docker-compose-web-with-gpu.yml up要在本地构建 docker 映像，您可以运行（您需要在您的计算机上进行 make）
make build-image然后测试构建的图像运行
make run-web-server # use `--use-gpu` for speedup if you have a compatible NVIDIA GPU.
# use `--target-lang <language_code>` to specify a target language.
# use `--inpainter=none` to disable inpainting.
# use `--translator=none` if you only want to use inpainting (blank bubbles)
# replace <path> with the path to the image folder or file.
$ python -m manga_translator -v --translator=google -l ENG -i < path >
# results can be found under `<path_to_image_folder>-translated`. # saves singular image into /result folder for demonstration purposes
# use `--mode demo` to enable demo translation.
# replace <path> with the path to the image file.
$ python -m manga_translator --mode demo -v --translator=google -l ENG -i < path >
# result can be found in `result/`. # use `--mode web` to start a web server.
$ python -m manga_translator -v --mode web --use-gpu
# the demo will be serving on http://127.0.0.1:5003 # use `--mode web` to start a web server.
$ python -m manga_translator -v --mode api --use-gpu
# the demo will be serving on http://127.0.0.1:5003 GUI 实现：BallonsTranslator
探测器：
--detector ctd可以增加检测到的文本行数量光学字符识别：
译者：
油漆工：??
着色器： mc2
--upscale-ratio 2或任何其他值来使用放大器--font-size-minimum 30或使用--manga2eng渲染器，该渲染器将尝试适应检测到的文本气泡--font-path fonts/anime_ace_3.ttf指定字体 -h, --help show this help message and exit
-m, --mode {demo,batch,web,web_client,ws,api}
Run demo in single image demo mode (demo), batch
translation mode (batch), web service mode (web)
-i, --input INPUT [INPUT ...] Path to an image file if using demo mode, or path to an
image folder if using batch mode
-o, --dest DEST Path to the destination folder for translated images in
batch mode
-l, --target-lang {CHS,CHT,CSY,NLD,ENG,FRA,DEU,HUN,ITA,JPN,KOR,PLK,PTB,ROM,RUS,ESP,TRK,UKR,VIN,ARA,CNR,SRP,HRV,THA,IND,FIL}
Destination language
-v, --verbose Print debug info and save intermediate images in result
folder
-f, --format {png,webp,jpg,xcf,psd,pdf} Output format of the translation.
--attempts ATTEMPTS Retry attempts on encountered error. -1 means infinite
times.
--ignore-errors Skip image on encountered error.
--overwrite Overwrite already translated images in batch mode.
--skip-no-text Skip image without text (Will not be saved).
--model-dir MODEL_DIR Model directory (by default ./models in project root)
--use-gpu Turn on/off gpu
--use-gpu-limited Turn on/off gpu (excluding offline translator)
--detector {default,ctd,craft,none} Text detector used for creating a text mask from an
image, DO NOT use craft for manga, it's not designed
for it
--ocr {32px,48px,48px_ctc,mocr} Optical character recognition (OCR) model to use
--use-mocr-merge Use bbox merge when Manga OCR inference.
--inpainter {default,lama_large,lama_mpe,sd,none,original}
Inpainting model to use
--upscaler {waifu2x,esrgan,4xultrasharp} Upscaler to use. --upscale-ratio has to be set for it
to take effect
--upscale-ratio UPSCALE_RATIO Image upscale ratio applied before detection. Can
improve text detection.
--colorizer {mc2} Colorization model to use.
--translator {google,youdao,baidu,deepl,papago,caiyun,gpt3,gpt3.5,gpt4,none,original,offline,nllb,nllb_big,sugoi,jparacrawl,jparacrawl_big,m2m100,m2m100_big,sakura}
Language translator to use
--translator-chain TRANSLATOR_CHAIN Output of one translator goes in another. Example:
--translator-chain "google:JPN;sugoi:ENG".
--selective-translation SELECTIVE_TRANSLATION
Select a translator based on detected language in
image. Note the first translation service acts as
default if the language isn't defined. Example:
--translator-chain "google:JPN;sugoi:ENG".
--revert-upscaling Downscales the previously upscaled image after
translation back to original size (Use with --upscale-
ratio).
--detection-size DETECTION_SIZE Size of image used for detection
--det-rotate Rotate the image for detection. Might improve
detection.
--det-auto-rotate Rotate the image for detection to prefer vertical
textlines. Might improve detection.
--det-invert Invert the image colors for detection. Might improve
detection.
--det-gamma-correct Applies gamma correction for detection. Might improve
detection.
--unclip-ratio UNCLIP_RATIO How much to extend text skeleton to form bounding box
--box-threshold BOX_THRESHOLD Threshold for bbox generation
--text-threshold TEXT_THRESHOLD Threshold for text detection
--min-text-length MIN_TEXT_LENGTH Minimum text length of a text region
--no-text-lang-skip Dont skip text that is seemingly already in the target
language.
--inpainting-size INPAINTING_SIZE Size of image used for inpainting (too large will
result in OOM)
--inpainting-precision {fp32,fp16,bf16} Inpainting precision for lama, use bf16 while you can.
--colorization-size COLORIZATION_SIZE Size of image used for colorization. Set to -1 to use
full image size
--denoise-sigma DENOISE_SIGMA Used by colorizer and affects color strength, range
from 0 to 255 (default 30). -1 turns it off.
--mask-dilation-offset MASK_DILATION_OFFSET By how much to extend the text mask to remove left-over
text pixels of the original image.
--font-size FONT_SIZE Use fixed font size for rendering
--font-size-offset FONT_SIZE_OFFSET Offset font size by a given amount, positive number
increase font size and vice versa
--font-size-minimum FONT_SIZE_MINIMUM Minimum output font size. Default is
image_sides_sum/200
--font-color FONT_COLOR Overwrite the text fg/bg color detected by the OCR
model. Use hex string without the "#" such as FFFFFF
for a white foreground or FFFFFF:000000 to also have a
black background around the text.
--line-spacing LINE_SPACING Line spacing is font_size * this value. Default is 0.01
for horizontal text and 0.2 for vertical.
--force-horizontal Force text to be rendered horizontally
--force-vertical Force text to be rendered vertically
--align-left Align rendered text left
--align-center Align rendered text centered
--align-right Align rendered text right
--uppercase Change text to uppercase
--lowercase Change text to lowercase
--no-hyphenation If renderer should be splitting up words using a hyphen
character (-)
--manga2eng Render english text translated from manga with some
additional typesetting. Ignores some other argument
options
--gpt-config GPT_CONFIG Path to GPT config file, more info in README
--use-mtpe Turn on/off machine translation post editing (MTPE) on
the command line (works only on linux right now)
--save-text Save extracted text and translations into a text file.
--save-text-file SAVE_TEXT_FILE Like --save-text but with a specified file path.
--filter-text FILTER_TEXT Filter regions by their text with a regex. Example
usage: --text-filter ".*badtext.*"
--pre-dict FILe_PATH Path to the pre-translation dictionary file. One entry per line,
Comments can be added with `#` and `//`.
usage: //Example
dog cat #Example
abc def
abc
--post-dict FILE_PATH Path to the post-translation dictionary file. Same as above.
--skip-lang Skip translation if source image is one of the provide languages,
use comma to separate multiple languages. Example: JPN,ENG
--prep-manual Prepare for manual typesetting by outputting blank,
inpainted images, plus copies of the original for
reference
--font-path FONT_PATH Path to font file
--gimp-font GIMP_FONT Font family to use for gimp rendering.
--host HOST Used by web module to decide which host to attach to
--port PORT Used by web module to decide which port to attach to
--nonce NONCE Used by web module as secret for securing internal web
server communication
--ws-url WS_URL Server URL for WebSocket mode
--save-quality SAVE_QUALITY Quality of saved JPEG image, range from 0 to 100 with
100 being best
--ignore-bubble IGNORE_BUBBLE The threshold for ignoring text in non bubble areas,
with valid values ranging from 1 to 50, does not ignore
others. Recommendation 5 to 10. If it is too low,
normal bubble areas may be ignored, and if it is too
large, non bubble areas may be considered normal
bubbles
由--target-lang或-l参数使用。
CHS : Chinese (Simplified)
CHT : Chinese (Traditional)
CSY : Czech
NLD : Dutch
ENG : English
FRA : French
DEU : German
HUN : Hungarian
ITA : Italian
JPN : Japanese
KOR : Korean
PLK : Polish
PTB : Portuguese (Brazil)
ROM : Romanian
RUS : Russian
ESP : Spanish
TRK : Turkish
UKR : Ukrainian
VIN : Vietnames
ARA : Arabic
SRP : Serbian
HRV : Croatian
THA : Thai
IND : Indonesian
FIL : Filipino (Tagalog)| 姓名 | API密钥 | 离线 | 笔记 |
|---|---|---|---|
| 暂时禁用 | |||
| 有道 | ✔️ | 需要YOUDAO_APP_KEY和YOUDAO_SECRET_KEY | |
| 百度 | ✔️ | 需要BAIDU_APP_ID和BAIDU_SECRET_KEY | |
| 深度 | ✔️ | 需要DEEPL_AUTH_KEY | |
| 彩云 | ✔️ | 需要CAIYUN_TOKEN | |
| 总蛋白三 | ✔️ | 实现text-davinci-003。需要OPENAI_API_KEY | |
| gpt3.5 | ✔️ | 实现 gpt-3.5-turbo。需要OPENAI_API_KEY | |
| 组蛋白4 | ✔️ | 实现 gpt-4。需要OPENAI_API_KEY | |
| 帕帕戈 | |||
| 樱花 | 需要SAKURA_API_BASE | ||
| 离线 | ✔️ | 选择最适合语言的离线翻译器 | |
| 苏戈伊 | ✔️ | Sugoi V4.0 型号 | |
| 米2米100 | ✔️ | 支持每种语言 | |
| m2m100_big | ✔️ | ||
| 没有任何 | ✔️ | 翻译为空文本 | |
| 原来的 | ✔️ | 保留原文 |
OPENAI_API_KEY = sk-xxxxxxx...
DEEPL_AUTH_KEY = xxxxxxxx...离线：翻译器是否可以离线使用。
Sugoi由mingshiba创建，请支持他https://www.patreon.com/mingshiba
由--gpt-config参数使用。
# The prompt being feed into GPT before the text to translate.
# Use {to_lang} to indicate where the target language name should be inserted.
# Note: ChatGPT models don't use this prompt.
prompt_template : >
Please help me to translate the following text from a manga to {to_lang}
(if it's already in {to_lang} or looks like gibberish you have to output it as it is instead):n
# What sampling temperature to use, between 0 and 2.
# Higher values like 0.8 will make the output more random,
# while lower values like 0.2 will make it more focused and deterministic.
temperature : 0.5
# An alternative to sampling with temperature, called nucleus sampling,
# where the model considers the results of the tokens with top_p probability mass.
# So 0.1 means only the tokens comprising the top 10% probability mass are considered.
top_p : 1
# The prompt being feed into ChatGPT before the text to translate.
# Use {to_lang} to indicate where the target language name should be inserted.
# Tokens used in this example: 57+
chat_system_template : >
You are a professional translation engine,
please translate the story into a colloquial,
elegant and fluent content,
without referencing machine translations.
You must only translate the story, never interpret it.
If there is any issue in the text, output it as is.
Translate to {to_lang}.
# Samples being feed into ChatGPT to show an example conversation.
# In a [prompt, response] format, keyed by the target language name.
#
# Generally, samples should include some examples of translation preferences, and ideally
# some names of characters it's likely to encounter.
#
# If you'd like to disable this feature, just set this to an empty list.
chat_sample :
Simplified Chinese : # Tokens used in this example: 88 + 84
- <|1|>恥ずかしい… 目立ちたくない… 私が消えたい…
<|2|>きみ… 大丈夫⁉
<|3|>なんだこいつ 空気読めて ないのか…？
- <|1|>好尴尬…我不想引人注目…我想消失…
<|2|>你…没事吧⁉
<|3|>这家伙怎么看不懂气氛的…？
# Overwrite configs for a specific model.
# For now the list is: gpt3, gpt35, gpt4
gpt35 :
temperature : 0.3当将输出格式设置为 { xcf , psd , pdf } 时，将使用 Gimp 生成文件。
在 Windows 上，假定 Gimp 2.x 安装到C:Users<Username>AppDataLocalProgramsGimp 2 。
生成的.xcf文件包含原始图像作为最低层，并将修复作为单独的层。翻译后的文本框有自己的图层，以原始文本作为图层名称，以便于访问。
限制：
.psd文件时，Gimp 会将文本图层转换为常规图像。--gimp-font参数单独控制的。 # use `--mode api` to start a web server.
$ python -m manga_translator -v --mode api --use-gpu
# the api will be serving on http://127.0.0.1:5003 API 接受 json(post) 和 multipart。
API 端点为/colorize_translate 、 /inpaint_translate 、 /translate 、 /get_text 。
api 的有效参数是：
// These are taken from args.py. For more info see README.md
detector: String
ocr: String
inpainter: String
upscaler: String
translator: String
target_language: String
upscale_ratio: Integer
translator_chain: String
selective_translation: String
attempts: Integer
detection_size: Integer // 1024 => 'S', 1536 => 'M', 2048 => 'L', 2560 => 'X'
text_threshold: Float
box_threshold: Float
unclip_ratio: Float
inpainting_size: Integer
det_rotate: Bool
det_auto_rotate: Bool
det_invert: Bool
det_gamma_correct: Bool
min_text_length: Integer
colorization_size: Integer
denoise_sigma: Integer
mask_dilation_offset: Integer
ignore_bubble: Integer
gpt_config: String
filter_text: String
overlay_type: String
// These are api specific args
direction: String // {'auto', 'h', 'v'}
base64Images: String //Image in base64 format
image: Multipart // image upload from multipart
url: String // an url string
手动翻译以人工翻译取代机器翻译。使用网页模式时，可以在 http://127.0.0.1:5003/manual 找到基本的手动翻译演示。
该demo提供了两种模式的翻译服务：同步模式和异步模式。
在同步模式下，一旦翻译任务完成，您的 HTTP POST 请求就会完成。
在异步模式下，您的 HTTP POST 请求将立即响应一个task_id ，您可以使用此task_id来轮询翻译任务状态。
file:<content-of-image>到 http://127.0.0.1:5003/runtask_id在result/目录中查找翻译结果，例如使用Nginx公开result/ file:<content-of-image>到 http://127.0.0.1:5003/submittask_id{"taskid": <task-id>}发布到 http://127.0.0.1:5003/task-state 来轮询翻译任务状态finished 、 error或error-lang时，翻译完成result/目录中查找翻译结果，例如使用 Nginx 公开result/ 将带有表单数据file:<content-of-image>到 http://127.0.0.1:5003/manual-translate 并等待响应。
您将获得如下 JSON 响应：
{
"task_id" : " 12c779c9431f954971cae720eb104499 " ,
"status" : " pending " ,
"trans_result" : [
{
"s" : " ☆上司来ちゃった…… " ,
"t" : " "
}
]
}填写翻译文本：
{
"task_id" : " 12c779c9431f954971cae720eb104499 " ,
"status" : " pending " ,
"trans_result" : [
{
"s" : " ☆上司来ちゃった…… " ,
"t" : " ☆Boss is here... "
}
]
}将翻译后的 JSON 发布到 http://127.0.0.1:5003/post-manual-result 并等待响应。
然后你可以在result/目录中找到翻译结果，例如使用 Nginx 公开result/ 。
接下来需要做什么的列表，欢迎您贡献。
GPU服务器并不便宜，请考虑捐赠给我们。
Ko-fi：https://ko-fi.com/voilelabs
帕特隆：https://www.patreon.com/voilelabs
爱发电：https://afdian.net/@voilelabs
