Mask-surveillance-Mask-detection-

Detects a person wearing a mask or not in real time and through image

In this Covid-19 Pandemic situation, this Deep learning model can be used as a surveillance system to detect a person wearing a Mask or Not in real time and through static image.

Dataset:

Create a Dataset with images of people wearing mask and without mask. Each having around 500 plus images.

Here there are two main files,

• training_mask.py : This file takes the input Dataset and trains on the images using VGG16 architecture and the created Model is saved in the respective directory.
• detect_mask.py : Used to detect face masks on the Static images using the saved model.

Training the Mask detector model:

• In the TrainingModel folder, there is Training_mask.ipnb and training_mask.py files. You can use either of the two files to train the model.
• Pass the directory having Mask and WithoutMask folder to training.
• For training I have used VGG16 architecture and imagenet weights.
• Save the Model

Face Mask detection:

• Use detect_mask.py file for Face mask detection.
• Faces in the image are detected first using Opencv and prediction is made on top of this detected image.
• ROI and detection label is created on the original image.