def resize(path_all, path_resize, mode):
image_names_dir = os.listdir(path_all)
if mode == 'Test_Frame':
image_names_dir = os.listdir(path_all + 'No_Fire')
image_names_dir.sort()
new_width = new_size.get('width')
new_height = new_size.get('height')
dimension = (new_width, new_height)
count = 0
for image in image_names_dir:
if mode == 'Fire':
img = cv2.imread(path_all + '/' + image)
resized_img = cv2.resize(img,
dimension,
interpolation=cv2.INTER_AREA)
cv2.imwrite(path_resize + '/resized_' + image, resized_img)
elif mode == 'Lake_Mary':
img = cv2.imread(path_all + '/' + image)
resized_img = cv2.resize(img,
dimension,
interpolation=cv2.INTER_AREA)
cv2.imwrite(path_resize + '/lake_resized_' + image, resized_img)
elif mode == 'Test_Frame':
img = cv2.imread(path_all + 'No_Fire/' + image)
resized_img = cv2.resize(img,
dimension,
interpolation=cv2.INTER_AREA)
cv2.imwrite(path_resize + 'No_Fire/resized_' + image, resized_img)
print('Image after Resizing ' + str(count) + ' : resized_' + image)
count += 1
def resize(path_all, path_resize, mode):
"""
Resizing the imported images to the project and save them on drive based on the dimension parameter.
:param path_all: The directory of loaded images to the project
:param path_resize: The directory to save the resized files
:param mode: Fire, No_Fire(lake mary), or the test data
:return: None
"""
image_names_dir = os.listdir(path_all)
if mode == 'Test_Frame':
# image_names_dir = os.listdir(path_all + 'Fire') # This is for the FIRE DIR (Test)
image_names_dir = os.listdir(
path_all + 'No_Fire') # This is for the No_FIRE DIR (Test)
image_names_dir.sort()
new_width = new_size.get('width')
new_height = new_size.get('height')
dimension = (new_width, new_height)
count = 0
for image in image_names_dir:
# print(resized_img.shape)
# cv2.imshow('output', resized_img)
if mode == 'Fire':
img = cv2.imread(path_all + '/' + image)
resized_img = cv2.resize(img,
dimension,
interpolation=cv2.INTER_AREA)
cv2.imwrite(path_resize + '/resized_' + image, resized_img)
elif mode == 'Lake_Mary':
img = cv2.imread(path_all + '/' + image)
resized_img = cv2.resize(img,
dimension,
interpolation=cv2.INTER_AREA)
cv2.imwrite(path_resize + '/lake_resized_' + image, resized_img)
elif mode == 'Test_Frame':
# img = cv2.imread(path_all + 'Fire/' + image)
img = cv2.imread(path_all + 'No_Fire/' + image)
resized_img = cv2.resize(img,
dimension,
interpolation=cv2.INTER_AREA)
# cv2.imwrite(path_resize + 'Fire/resized_' + image, resized_img) # Resize for Fire (Test Data)
cv2.imwrite(path_resize + 'No_Fire/resized_' + image,
resized_img) # Resize for NoFire (Test Data)
print('Image Resized ' + str(count) + ' : resized_' + image)
count += 1
def resize(path_all, path_resize, mode): #Function to resize
image_names_dir = os.listdir(path_all)
if mode == 'Test_Frame':
# image_names_dir = os.listdir(path_all + 'Fire') # This is for the FIRE DIR (Test)
image_names_dir = os.listdir(
path_all + 'No_Fire') # This is for the No_FIRE DIR (Test)
image_names_dir.sort()
new_width = new_size.get('width')
new_height = new_size.get('height')
dimension = (new_width, new_height)
count = 0
for image in image_names_dir:
# print(resized_img.shape)
# cv2.imshow('output', resized_img)
if mode == 'Fire':
img = cv2.imread(path_all + '/' + image)
resized_img = cv2.resize(img,
dimension,
interpolation=cv2.INTER_AREA)
cv2.imwrite(path_resize + '/resized_' + image, resized_img)
elif mode == 'Lake_Mary':
img = cv2.imread(path_all + '/' + image)
resized_img = cv2.resize(img,
dimension,
interpolation=cv2.INTER_AREA)
cv2.imwrite(path_resize + '/lake_resized_' + image, resized_img)
elif mode == 'Test_Frame':
# img = cv2.imread(path_all + 'Fire/' + image)
img = cv2.imread(path_all + 'No_Fire/' + image)
resized_img = cv2.resize(img,
dimension,
interpolation=cv2.INTER_AREA)
# cv2.imwrite(path_resize + 'Fire/resized_' + image, resized_img) # Resize for Fire (Test Data)
cv2.imwrite(path_resize + 'No_Fire/resized_' + image,
resized_img) # Resize for NoFire (Test Data)
print('Image Resized ' + str(count) + ' : resized_' + image)
count += 1
OS: Ubuntu 18.04
################################
"""
#########################################################
# import libraries
import numpy as np
import tensorflow as tf
from tensorflow.keras.models import load_model
from plotdata import plot_confusion_matrix
from config import Config_classification
from config import new_size
batch_size = Config_classification.get('batch_size')
image_size = (new_size.get('width'), new_size.get('height'))
epochs = Config_classification.get('Epochs')
#########################################################
# Function definition
def classify():
"""
This function load the trained model from the previous task and evaluates the performance of that over the test
data set.
:return: None, Plot the Confusion matrix for the test data on the binary classification
"""
test_ds = tf.keras.preprocessing.image_dataset_from_directory(
"frames/Test",
seed=1337,