A Python API wrapper for removing backgrounds from picture using remove.bg's API.
This code is licensed under the MIT License. See here for more details.
pip install removebg
Removes the background given an image file.
Parameter | Default Value | Description |
---|---|---|
img_file_path | req. param | path to the source image file |
size | 'regular' |
size of the output image ('auto' = highest available resolution, 'preview' |
type | 'auto' |
foreground object ('auto' = autodetect, 'person' , 'product' , 'car' ) |
type_level | 'none' |
classification level of the foreground object ('none' = no classification, '1' = coarse classification (e.g. 'car' ), '2' = specific classification (e.g. 'car_interior' ), 'latest' = latest classification) |
format | 'auto' |
image format ('auto' = autodetect, 'png' , 'jpg' , 'zip' ) |
roi | '0 0 100% 100%' |
region of interest, where to look for foreground object (x1, y1, x2, y2) in px or relative (%) |
crop | None |
px or relative, single val = all sides, two vals = top/bottom, left/right, four vals = top, right, bottom, left |
scale | 'original' |
image scale relative to the total image size |
position | 'original' |
'center' , 'original' , single val = horizontal and vertical, two vals = horizontal, vertical |
channels | 'rgba' |
request the finalized image ('rgba' ) or an alpha mask ('alpha' ) |
shadow | False |
whether to add an artificial shadow (some types aren't supported) |
semitransparency | True |
semitransparency for windows or glass objects (some types aren't supported) |
bg | None |
background (None = no background, path, url, color hex code (e.g. '81d4fa' , 'fff' ), color name (e.g. 'green' )) |
bg_type | None |
background type (None = no background, 'path' , 'url' , 'color' ) |
new_file_name | 'no-bg.png' |
file name of the result image |
from removebg import RemoveBg
rmbg = RemoveBg("YOUR-API-KEY", "error.log")
rmbg.remove_background_from_img_file("joker.jpg")
Removes the background given an image URL.
Parameter | Default Value | Description |
---|---|---|
img_url | req. param | URL to the source image |
size | 'regular' |
size of the output image ('auto' = highest available resolution, 'preview' |
type | 'auto' |
foreground object ('auto' = autodetect, 'person' , 'product' , 'car' ) |
type_level | 'none' |
classification level of the foreground object ('none' = no classification, '1' = coarse classification (e.g. 'car' ), '2' = specific classification (e.g. 'car_interior' ), 'latest' = latest classification) |
format | 'auto' |
image format ('auto' = autodetect, 'png' , 'jpg' , 'zip' ) |
roi | '0 0 100% 100%' |
region of interest, where to look for foreground object (x1, y1, x2, y2) in px or relative (%) |
crop | None |
px or relative, single val = all sides, two vals = top/bottom, left/right, four vals = top, right, bottom, left |
scale | 'original' |
image scale relative to the total image size |
position | 'original' |
'center' , 'original' , single val = horizontal and vertical, two vals = horizontal, vertical |
channels | 'rgba' |
request the finalized image ('rgba' ) or an alpha mask ('alpha' ) |
shadow | False |
whether to add an artificial shadow (some types aren't supported) |
semitransparency | True |
semitransparency for windows or glass objects (some types aren't supported) |
bg | None |
background (None = no background, path, url, color hex code (e.g. '81d4fa' , 'fff' ), color name (e.g. 'green' )) |
bg_type | None |
background type (None = no background, 'path' , 'url' , 'color' ) |
new_file_name | 'no-bg.png' |
file name of the result image |
from removebg import RemoveBg
rmbg = RemoveBg("YOUR-API-KEY", "error.log")
rmbg.remove_background_from_img_url("http://www.example.com/some_image.jpg")
Removes the background given a base64 image string.
Parameter | Default Value | Description |
---|---|---|
base64_img | req. param | base64 image string |
size | 'regular' |
size of the output image ('auto' = highest available resolution, 'preview' |
type | 'auto' |
foreground object ('auto' = autodetect, 'person' , 'product' , 'car' ) |
type_level | 'none' |
classification level of the foreground object ('none' = no classification, '1' = coarse classification (e.g. 'car' ), '2' = specific classification (e.g. 'car_interior' ), 'latest' = latest classification) |
format | 'auto' |
image format ('auto' = autodetect, 'png' , 'jpg' , 'zip' ) |
roi | '0 0 100% 100%' |
region of interest, where to look for foreground object (x1, y1, x2, y2) in px or relative (%) |
crop | None |
px or relative, single val = all sides, two vals = top/bottom, left/right, four vals = top, right, bottom, left |
scale | 'original' |
image scale relative to the total image size |
position | 'original' |
'center' , 'original' , single val = horizontal and vertical, two vals = horizontal, vertical |
channels | 'rgba' |
request the finalized image ('rgba' ) or an alpha mask ('alpha' ) |
shadow | False |
whether to add an artificial shadow (some types aren't supported) |
semitransparency | True |
semitransparency for windows or glass objects (some types aren't supported) |
bg | None |
background (None = no background, path, url, color hex code (e.g. '81d4fa' , 'fff' ), color name (e.g. 'green' )) |
bg_type | None |
background type (None = no background, 'path' , 'url' , 'color' ) |
new_file_name | 'no-bg.png' |
file name of the result image |
from removebg import RemoveBg
import base64
rmbg = RemoveBg("YOUR-API-KEY", "error.log")
with open("joker.jpg", "rb") as image_file:
encoded_string = base64.b64encode(image_file.read())
rmbg.remove_background_from_base64_img(encoded_string)
Contributions and feature requests are always welcome.