# Import necessary libraries
import cv2
import numpy as np
import matplotlib.pyplot as plt
from skimage import measure, morphology, exposure
from scipy import ndimage
# Install necessary library for file upload in Google Colab
try:
from google.colab import files
except ImportError:
pass
# Function to detect tumor and visualize results
def detect_and_visualize_tumor(image_path):
image = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
# Check if the image was loaded properly
if image is None:
print(f"Error: Unable to load the image {image_path}. Please check the
image path.")
return
print(f"Image {image_path} loaded successfully.")
# Display the original image
plt.figure()
plt.imshow(image, cmap='gray')
plt.title('Original Image')
plt.axis('off')
plt.show()
# Apply median filter to the original image
image_filtered = cv2.medianBlur(image, 5)
# Display the filtered image
plt.figure()
plt.imshow(image_filtered, cmap='gray')
plt.title('Filtered Image')
plt.axis('off')
plt.show()
# Enhance the filtered image using histogram equalization
image_enhanced = exposure.equalize_hist(image_filtered)
# Apply Gaussian blur to reduce noise
image_blurred = cv2.GaussianBlur(image_enhanced, (5, 5), 0)
# Edge detection using Canny
edges = cv2.Canny((image_blurred * 255).astype(np.uint8), 100, 200)
# Display the enhanced image
plt.figure()
plt.imshow(image_blurred, cmap='gray')
plt.title('Enhanced Image')
plt.axis('off')
plt.show()
# Display edge detection
plt.figure()
plt.imshow(edges, cmap='gray')
plt.title('Edge Detection')
plt.axis('off')
plt.show()
# Tumor detection function using Otsu's thresholding
def detect_tumor(image):
# Apply Otsu's thresholding
_, bw = cv2.threshold((image * 255).astype(np.uint8), 0, 255,
cv2.THRESH_BINARY + cv2.THRESH_OTSU)
label_img = measure.label(bw)
props = measure.regionprops(label_img)
# Consider regions with solidity > 0.7 and significant area
high_dense_area = [prop for prop in props if prop.solidity > 0.7 and
prop.area > 100] # Adding area constraint
if not high_dense_area:
return None, None
max_area = max(high_dense_area, key=lambda x: x.area)
tumor_mask = label_img == max_area.label
tumor_mask = morphology.dilation(tumor_mask, morphology.square(5))
return tumor_mask, max_area.bbox
# Detect the tumor
tumor_mask, bbox = detect_tumor(image_blurred)
if tumor_mask is None:
print("No tumor detected.")
return
# Find contours around the tumor
contours, _ = cv2.findContours((tumor_mask * 255).astype(np.uint8),
cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
# Draw contours on the original image
image_with_contours = cv2.cvtColor(image, cv2.COLOR_GRAY2BGR)
cv2.drawContours(image_with_contours, contours, -1, (0, 0, 255), 2)
# Display the original image with tumor contours
plt.figure()
plt.imshow(image_with_contours)
plt.title('Detected Tumor with Contours')
plt.axis('off')
plt.show()
# Calculate and display the tumor area
tumor_area = np.sum(tumor_mask)
print(f"Tumor Area (in pixels): {tumor_area}")
# Crop and display the tumor region
min_row, min_col, max_row, max_col = bbox
cropped_tumor = image[min_row:max_row, min_col:max_col]
plt.figure()
plt.imshow(cropped_tumor, cmap='gray')
plt.title('Cropped Tumor Region')
plt.axis('off')
plt.show()
# Zoom into the cropped tumor region
zoom_factor = 2 # Define the zoom factor
zoomed_tumor = cv2.resize(cropped_tumor, None, fx=zoom_factor, fy=zoom_factor,
interpolation=cv2.INTER_LINEAR)
# Display the zoomed tumor image in a separate figure
plt.figure()
plt.imshow(zoomed_tumor, cmap='gray')
plt.title('Zoomed Tumor Region')
plt.axis('off')
plt.show()
# File upload in Google Colab
uploaded = files.upload()
# Process each uploaded image
for image_path in uploaded.keys():
detect_and_visualize_tumor(image_path)
