Scribd
Upload
13 views14 pages

Face Mask Detection Project Code

The project report details the implementation of a Face Mask Detection system using various algorithms, including CNN and VGG19. The report includes code for data preprocessing, model training, and evaluation metrics such as accuracy and loss. It highlights the dataset used, the architecture of the models, and visualizations of the results.

Uploaded by

ddiya.2610
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
13 views14 pages

Face Mask Detection Project Code

The project report details the implementation of a Face Mask Detection system using various algorithms, including CNN and VGG19. The report includes code for data preprocessing, model training, and evaluation metrics such as accuracy and loss. It highlights the dataset used, the architecture of the models, and visualizations of the results.

Uploaded by

ddiya.2610
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

School of Computer Science & Engineering

BCSE306L –Artificial Intelligence– C1 Slot – WIN 2023 -24

Project Report Source Code


DA-2

Project Title : Face Mask Detection using different algorithms


Review Number : II
Date of Submission : 27th March 2024

Members
1. Diya Das (21BCE5081)
2. Armaan Saini (21BCE5115)

Faculty In-charge:
Dr. Abirami S

Professor

SCOPE
Code for the normal CNN layers Implemented manually

# %%
## Importing libraries
import pandas as pd
import numpy as np
import cv2
import os
import glob
from [Link] import distance
from sklearn.model_selection import train_test_split
from [Link] import classification_report, confusion_matrix

import tensorflow as tf
from [Link] import Sequential, models
from [Link] import Flatten, Dense, Conv2D, MaxPool2D
from [Link] import ImageDataGenerator

import [Link] as plt


import seaborn as sns

# %%
path = "kaggle/input/face-mask-12k-images-dataset/Face Mask Dataset/"

# %%
dataset = {
"image_path": [],
"mask_status": [],
"where": []
}

for where in [Link](path):


for status in [Link](path+"/"+where):
for image in [Link](path+where+"/"+status+"/"+"*.png"):
dataset["image_path"].append(image)
dataset["mask_status"].append(status)
dataset["where"].append(where)

dataset = [Link](dataset)
[Link]()

# %%
## Choosing a random image to detect the face in the image
face_model =
[Link]('../input/haarcascades/haarcascade_frontalface_default.x
ml')

## Choosing the image from the directory


img = [Link]("../input/face-mask-detection/images/[Link]")
## Converting the image to grayscale to apply haarcascade algorithm
img = [Link](img, cv2.IMREAD_GRAYSCALE)

## Returns the x, y, w, h co-ordinates as numpy arrays for all the detected


faces
detected_face = face_model.detectMultiScale(img)

## Converting from grayscale to colored image


output_img = [Link](img, cv2.COLOR_RGB2BGR)

## Drawing rectangle box around the faces in the image


for (x, y, w, h) in detected_face:
[Link](output_img, (x,y), (x+w, y+h), (0, 0, 200), 2)

## Displaying the image


[Link](figsize = (15, 15))
[Link](output_img)

# %%
## Checking for total number of images in the dataset
print(f"With Mask:", dataset.value_counts("mask_status")[0])
print(f"Without Mask:", dataset.value_counts("mask_status")[1])

## Plotting the numbers


[Link](x = dataset["mask_status"])

# %%
[Link](figsize = (15, 10))

for i in range(9):
random = [Link](1, len(dataset))
[Link](3, 3, i+1)
[Link]([Link]([Link][random,"image_path"]))
[Link]([Link][random,"mask_status"], size = 15)
[Link]([])
[Link]([])

[Link]()

# %%
## Splitting train test and Validation Dataset
train_df = dataset[dataset["where"] == "Train"]
test_df = dataset[dataset["where"] == "Test"]
valid_df = dataset[dataset["where"] == "Validation"]

print(train_df.head(10))
## Shuffling the dataset
train_df = train_df.sample(frac = 1)
test_df = test_df.sample(frac = 1)
valid_df = valid_df.sample(frac = 1)

print("\n After Shuffling \n")


print(train_df.head(10))

# %%
[Link](figsize = (15, 5))
[Link](1, 3, 1)
[Link](x = train_df["mask_status"])
[Link]("Training Dataset", size = 10)

[Link](1, 3, 2)
[Link](x = test_df["mask_status"])
[Link]("Test Dataset", size = 10)

[Link](1, 3, 3)
[Link](x = valid_df["mask_status"])
[Link]("Validation Dataset", size = 10)

[Link]()

# %%
train_df = train_df.reset_index().drop("index", axis = 1)
train_df.head()

# %%
## Reading all the image into a list and changing the size of the image to
(150,150)
data = []
image_size = 150

for i in range(len(train_df)):
## Converting the image into grayscale
img_array = [Link](train_df["image_path"][i], cv2.IMREAD_GRAYSCALE)

## Resizing the array


new_image_array = [Link](img_array, (image_size, image_size))

##Encoding the image with the label


if train_df["mask_status"][i] == "WithMask":
[Link]([new_image_array, 1])
else:
[Link]([new_image_array, 0])

# %%
data = [Link](data,dtype=object)

# %%
data[0][0].shape

# %%
## Shuffling the data to make sure everything is not in order
[Link](data)

# %%
## Looking at the training samples

fig, ax = [Link](2, 3, figsize=(10, 10))

for row in range(2):


for col in range(3):
image_index = row*100+col

ax[row, col].axis("off")
ax[row,col].imshow(data[image_index][0], cmap = "gray")

if data[image_index][1] == 0:
ax[row, col].set_title("Without Mask")
else:
ax[row, col].set_title("With Mask")

[Link]()

# %%
X = []
y = []

## Seperating X and y
for image in data:
[Link](image[0])
[Link](image[1])

## Converting X and y to numpy array as Tensorflow accepts only numpy arrays


X = [Link](X)
y = [Link](y)

# %%
### Normalizing the data
X = X/255

### Train Test Split


X_train, X_val, y_train, y_val = train_test_split(X, y, test_size = 0.2,
random_state = 42)
# %%
model = Sequential()
[Link](Conv2D(64, (3, 3), activation = "relu"))
[Link](Conv2D(64, (3, 3), activation = "relu"))
[Link](MaxPool2D(pool_size=(3, 3)))
[Link](Flatten())
[Link](Dense(128, activation = "relu"))
[Link](Dense(1, activation = "sigmoid"))
# [Link]()

# %%
[Link](optimizer='adam',
loss=[Link](),
metrics=['accuracy'])

# %%
# X_train = X_train.reshape(-1, 32, 150, 150)
## Reshaping training set to match Conc2D
X_train = X_train.reshape(len(X_train), X_train.shape[1], X_train.shape[2], 1)
X_val = X_val.reshape(len(X_val), X_val.shape[1], X_val.shape[2], 1)

history = [Link](X_train, y_train, epochs=5, batch_size = 32)

# %%
[Link]()

# %%
def visualize_data(history):
# Plotting accuracy
[Link]([Link]['accuracy'])
[Link]('Model Accuracy')
[Link]('Epoch')
[Link]('Accuracy')
[Link](['Train', 'Validation'], loc='upper left')
[Link]()

# Plotting loss
[Link]([Link]['loss'])
[Link]('Model Loss')
[Link]('Epoch')
[Link]('Loss')
[Link](['Train', 'Validation'], loc='upper left')
[Link]()
visualize_data(history)

# %%
prediction = ([Link](X_val) > 0.5).astype("int32")
# %%
print(classification_report(y_val, prediction))
print(confusion_matrix(y_val, prediction))
Code for the VGG algorithm implemented.

# %%
# %%
## Importing libraries
import pandas as pd
import numpy as np
import cv2
import os
import glob
from [Link] import distance
from sklearn.model_selection import train_test_split
from [Link] import classification_report, confusion_matrix
from [Link].vgg19 import VGG19
import tensorflow as tf
from [Link] import Sequential, models
from [Link] import Flatten, Dense, Conv2D, MaxPool2D
from [Link] import ImageDataGenerator

import [Link] as plt


import seaborn as sns

# %%
# %%
path = "/kaggle/input/face-mask-12k-images-dataset/Face Mask Dataset/"

# %%
# %%
dataset = {
"image_path": [],
"mask_status": [],
"where": []
}

for where in [Link](path):


for status in [Link](path+"/"+where):
for image in [Link](path+where+"/"+status+"/"+"*.png"):
dataset["image_path"].append(image)
dataset["mask_status"].append(status)
dataset["where"].append(where)

dataset = [Link](dataset)
[Link]()

# %%

# %%
## Choosing a random image to detect the face in the image
face_model =
[Link]('/kaggle/input/haarcascade/haarcascade_frontalface_defau
[Link]')

## Choosing the image from the directory


img = [Link]("/kaggle/input/face-mask-
detection/images/[Link]")

## Converting the image to grayscale to apply haarcascade algorithm


img = [Link](img, cv2.IMREAD_GRAYSCALE)

## Returns the x, y, w, h co-ordinates as numpy arrays for all the detected


faces
detected_face = face_model.detectMultiScale(img)

## Converting from grayscale to colored image


output_img = [Link](img, cv2.COLOR_RGB2BGR)

## Drawing rectangle box around the faces in the image


for (x, y, w, h) in detected_face:
[Link](output_img, (x,y), (x+w, y+h), (0, 0, 200), 2)

## Displaying the image


[Link](figsize = (15, 15))
[Link](output_img)

# %%

# %%
## Checking for total number of images in the dataset
print(f"With Mask:", dataset.value_counts("mask_status")[0])
print(f"Without Mask:", dataset.value_counts("mask_status")[1])

## Plotting the numbers


[Link](x = dataset["mask_status"])

# %%

# %%
[Link](figsize = (15, 10))

for i in range(9):
random = [Link](1, len(dataset))
[Link](3, 3, i+1)
[Link]([Link]([Link][random,"image_path"]))
[Link]([Link][random,"mask_status"], size = 15)
[Link]([])
[Link]([])
[Link]()

# %%

# %%
## Splitting train test and Validation Dataset
train_df = dataset[dataset["where"] == "Train"]
test_df = dataset[dataset["where"] == "Test"]
valid_df = dataset[dataset["where"] == "Validation"]

print(train_df.head(10))

## Shuffling the dataset


train_df = train_df.sample(frac = 1)
test_df = test_df.sample(frac = 1)
valid_df = valid_df.sample(frac = 1)

print("\n After Shuffling \n")


print(train_df.head(10))

# %%

# %%
[Link](figsize = (15, 5))
[Link](1, 3, 1)
[Link](x = train_df["mask_status"])
[Link]("Training Dataset", size = 10)

[Link](1, 3, 2)
[Link](x = test_df["mask_status"])
[Link]("Test Dataset", size = 10)

[Link](1, 3, 3)
[Link](x = valid_df["mask_status"])
[Link]("Validation Dataset", size = 10)

[Link]()

# %%
# %%
train_df = train_df.reset_index().drop("index", axis = 1)
train_df.head()

# %%

# %%
## Reading all the image into a list and changing the size of the image to
(150,150)
data = []
image_size = 150

for i in range(len(train_df)):
## Converting the image into grayscale
img_array = [Link](train_df["image_path"][i], cv2.IMREAD_GRAYSCALE)

## Resizing the array


new_image_array = [Link](img_array, (image_size, image_size))

##Encoding the image with the label


if train_df["mask_status"][i] == "WithMask":
[Link]([new_image_array, 1])
else:
[Link]([new_image_array, 0])

# %%

# %%
data = [Link](data,dtype="object")

# %%

# %%
data[0][0].shape

# %%

# %%
## Shuffling the data to make sure everything is not in order
[Link](data)

# %%

# %%
## Looking at the training samples

fig, ax = [Link](2, 3, figsize=(10, 10))

for row in range(2):


for col in range(3):
image_index = row*100+col

ax[row, col].axis("off")
ax[row,col].imshow(data[image_index][0], cmap = "gray")
if data[image_index][1] == 0:
ax[row, col].set_title("Without Mask")
else:
ax[row, col].set_title("With Mask")

[Link]()

# %%

# %%
# datagen = ImageDataGenerator(rescale = 1./255)

# train_generator=datagen.flow_from_dataframe(
# dataframe=train_df,
# directory="../input",
# x_col="image_path",
# y_col="mask_status",
# batch_size=80,
# seed=42,
# shuffle=False,
# class_mode="binary",
# target_size=(150,150))

# %%

# %%
X = []
y = []

## Seperating X and y
for image in data:
[Link](image[0])
[Link](image[1])

## Converting X and y to numpy array as Tensorflow accepts only numpy arrays


X = [Link](X)
y = [Link](y)

# %%

# %%
### Normalizing the data
X = X/255

### Train Test Split


X_train, X_val, y_train, y_val = train_test_split(X, y, test_size = 0.2,
random_state = 42)
# %%

# %%
vgg19 = VGG19(weights='imagenet',include_top=False,input_shape=(128, 128, 3))

for layer in [Link]:


[Link] = False

model = Sequential()
[Link](vgg19)
[Link](Flatten(name="layer1"))
[Link](Dense(128, activation='relu',name="layer2"))
[Link](Dense(2,activation='sigmoid',name="layer3"))
[Link]()

# %%
# %%
[Link](optimizer='adam',
loss=[Link](),
metrics=['accuracy'])

# %%

# %%
train_dir = '../input/face-mask-12k-images-dataset/Face Mask Dataset/Train'
test_dir = '../input/face-mask-12k-images-dataset/Face Mask Dataset/Test'
val_dir = '../input/face-mask-12k-images-dataset/Face Mask Dataset/Validation'

# %%
train_datagen = ImageDataGenerator(rescale=1.0/255, horizontal_flip=True,
zoom_range=0.2,shear_range=0.2)
train_generator =
train_datagen.flow_from_directory(directory=train_dir,target_size=(128,128),cl
ass_mode='categorical',batch_size=32)

val_datagen = ImageDataGenerator(rescale=1.0/255)
val_generator =
train_datagen.flow_from_directory(directory=val_dir,target_size=(128,128),clas
s_mode='categorical',batch_size=32)

test_datagen = ImageDataGenerator(rescale=1.0/255)
test_generator =
train_datagen.flow_from_directory(directory=val_dir,target_size=(128,128),clas
s_mode='categorical',batch_size=32)

history = [Link](train_generator,
epochs=5,
validation_data=val_generator)
# %%

# %%
def visualize_data(history):
# Plotting accuracy
[Link]([Link]['accuracy'])
[Link]('Model Accuracy')
[Link]('Epoch')
[Link]('Accuracy')
[Link](['Train', 'Validation'], loc='upper left')
[Link]()

# Plotting loss
[Link]([Link]['loss'])
[Link]('Model Loss')
[Link]('Epoch')
[Link]('Loss')
[Link](['Train', 'Validation'], loc='upper left')
[Link]()
visualize_data(history)

# %%
[Link]()

# %%

# %%
# Evaluate the model on the test data
evaluation = [Link](test_generator)

# Print the evaluation results


print("Test Loss:", evaluation[0])
print("Test Accuracy:", evaluation[1])

You might also like