takecx/remove_bg

remove background from image

remove_bg

Remove background from image

original image from PAKUTASO

Workflow

1. Create mask image using Image Segmentation technique.
2. Create trimap image using Morphology(cv2.erode(),cv2.dilate()).
3. Apply Image Matting using FBA_Matting

Supported foreground object type

Supported foreground object type depends on DEEPLABV3-RESNET101 output.
This model trained on PASCAL VOC 2012 semantic segmentation benchmark.
Class index of the dataset is shown as below.

[Class List]
1. aeroplane
2. bicycle
3. bird
4. boat
5. bottle
6. bus
7. car
8. cat
9. chair
10. cow
11. diningtable
12. dog
13. horse
14. motorbike
15. person
16. potted plant
17. sheep
18. sofa
19. train
20. tv/monitor

Supported input image format

• .png
• .jpg

How to Use

1. Clone this repo

$ git clone https://github.com/takecx/remove_bg.git
$ cd remove_bg

2. Restore submodule

$ git submodule update -i

3. Download FBA_Matting model.

Download FBA_Matting model from here and place it to ./FBA_Matting/FBA.pth

4. Create Environment

You should create python environment using requirements.txt.

$ conda create -n remove_bg python=3.7
$ conda activate remove_bg
$ pip install -r requirements.txt

5. Run script

Run BackgroundRemover.py with some arguments.

$ python BackgroundRemover.py (image_file) --kernel_size (morphology kernel size) --iteration (morphology iteration num)

Arguments

1. input image file path (required)
2. --kernel_size : kernel size of morphology conversion (optional)
3. --iteration : iteration num of morphology conversion (optional)

For example,

• input image file path : ./examples/images/sample.png
• --kernel_size : 5
• --iteration : 3

then, you run

$ python BackgroundRemover.py ./examples/images/sample.png --kernel_size 5 --iteration 3