Skip to content

Web App to identify Humans and extract Human Bodies from the image without Background

Notifications You must be signed in to change notification settings

KD1994/Body-Detection-With-BG-Removal

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Body-Detection-With-BG-Removal

This Web App is used to detect human bodies from the image and then extract the human bodies and apply BG removal technique to get just a human body without any kind of background.

Prerequisite

tensorflow
numpy  
scipy
opencv-python
pillow
matplotlib
h5py
keras
Flask

Installation

you can install all the above mentioned requirements using pip3.

1: imageai - Python library developed for object detection.

Refer the installation process here ---> here

All the prerequisites are mentioned above please install it before installing imageai.

use this command for installation of imageai:

pip3 install imageai --upgrade

I have used YoloV3 for object detection, you can use Retinanet or yolo-tiny which are also trained for it.

To download yolov3/Retinanet/yolo-tiny: here

2: Xception Model: used to remove the background of the image

use this command to download Xception model:

wget http://download.tensorflow.org/models/deeplabv3_pascal_train_aug_2018_01_04.tar.gz

Make sure to add these models into "static/models/(your_model_folder)" folder to run the main script.

Output Example

Body Detection Image with BG Image w/o BG
image image image
image image image
image image image
image image image
image image image

About

Web App to identify Humans and extract Human Bodies from the image without Background

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published