import cv2
import numpy as np
import matplotlib.pyplot as plt
import scipy.ndimage as sc

# --- 1. Citire si dimensiuni ---
img = cv2.imread('img1.jpg')
print(f"Dimensiuni: {img.shape}") 
if len(img.shape) == 3:
    print("Tip imagine: Color (BGR)")
else:
    print("Tip imagine: Grayscale")

# --- 2. Planul Rosu colorat in Cyan ---
# In OpenCV canalele sunt B-G-R. Rosu este indexul 2.
# Cyan = Albastru + Verde.
red_channel = img[:, :, 2]
img_cyan = np.zeros_like(img)
img_cyan[:, :, 0] = red_channel  # Canalul Blue primeste intensitatea de pe Red
img_cyan[:, :, 1] = red_channel  # Canalul Green primeste intensitatea de pe Red
# Canalul Red (2) ramane 0.

# --- 3. Transformare si Contrast (CLAHE) ---
# [cite_start]Se foloseste CLAHE pentru evidentierea detaliilor locale (tumora) [cite: 36]
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8, 8))
img_clahe = clahe.apply(gray)

# --- 4. Binarizare (Otsu) ---
# [cite_start]Extragerea tumorii folosind pragul Otsu [cite: 40]
ret, thresh = cv2.threshold(img_clahe, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
print(f"Pragul Otsu ales automat: {ret}")

# --- 5. Operatie Morfologica (Deschidere/Opening) ---
# [cite_start]Eliminare componente conexe mici fara a deteriora obiectul [cite: 71]
structura = np.ones((5, 5))
mask_clean = sc.binary_opening(thresh, structure=structura).astype(np.uint8) * 255

# --- Afisare ---
plt.figure(figsize=(12, 6))
plt.subplot(1, 3, 1); plt.imshow(cv2.cvtColor(img_cyan, cv2.COLOR_BGR2RGB)); plt.title('Plan Rosu -> Cyan')
plt.subplot(1, 3, 2); plt.imshow(img_clahe, cmap='gray'); plt.title('Contrast Imbunatatit (CLAHE)')
plt.subplot(1, 3, 3); plt.imshow(mask_clean, cmap='gray'); plt.title('Segmentare Finala (Tumora)')
plt.show()

import cv2
import numpy as np
import scipy.ndimage as sc
import matplotlib.pyplot as plt

img2 = cv2.imread('img2.jpg', 0)

# --- 1 & 2. Identificare si eliminare zgomot ---
# [cite_start]Daca zgomotul este Sare & Piper (impulsiv), se foloseste filtrul Median [cite: 69]
# Daca zgomotul este Gaussian, se poate folosi un filtru de mediere sau Gaussian.
# Presupunem filtru Median conform cerintelor uzuale de eliminare zgomot impulsiv:
img_filtered = sc.median_filter(img2, size=3, mode='nearest')

# --- 3. Imbunatatire contrast ---
# [cite_start]Egalizare histograma [cite: 35]
img_eq = cv2.equalizeHist(img_filtered)

# --- 4. Segmentare (Otsu) ---
# [cite_start]Binarizare pentru obtinerea obiectului [cite: 40]
ret2, mask2 = cv2.threshold(img_eq, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)

# --- 5. Extragere Zona de Interes (ROI) ---
# Aplicarea mastii peste imaginea filtrata
roi = cv2.bitwise_and(img_filtered, img_filtered, mask=mask2)

# --- Afisare ---
plt.figure(figsize=(12, 4))
plt.subplot(1, 4, 1); plt.imshow(img2, cmap='gray'); plt.title('Imagine Zgomotoasa')
plt.subplot(1, 4, 2); plt.imshow(img_filtered, cmap='gray'); plt.title('Filtrata (Median)')
plt.subplot(1, 4, 3); plt.imshow(mask2, cmap='gray'); plt.title('Masca (Otsu)')
plt.subplot(1, 4, 4); plt.imshow(roi, cmap='gray'); plt.title('Zona de Interes')
plt.show()

import cv2
import numpy as np
import matplotlib.pyplot as plt
from skimage import measure

img3 = cv2.imread('img3.jpg', 0)

# --- 1. Extragere Contur (Canny) ---
# [cite_start]Operatorul Canny pentru contururi [cite: 69]
edges = cv2.Canny(img3, 100, 200)

# --- 2. Pregatire pentru Arie si Perimetru ---
# Pentru a calcula aria, avem nevoie de regiuni pline, nu doar contururi.
# Aplicam binarizare si etichetare.
ret3, thresh3 = cv2.threshold(img3, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
[cite_start]labels = measure.label(thresh3) # Etichetare regiuni [cite: 74]

# [cite_start]Extragere proprietati [cite: 75]
props = measure.regionprops(labels, img3)

# [cite_start]Selectarea celulei albe (presupunem ca este regiunea cu aria cea mai mare) [cite: 32, 34]
celula = max(props, key=lambda r: r.area)

# --- 3. Afisare Arie si Perimetru ---
print(f"Aria celulei: {celula.area}")
print(f"Perimetrul celulei: {celula.perimeter}")

# --- Afisare ---
plt.figure(figsize=(10, 5))
plt.subplot(1, 2, 1); plt.imshow(edges, cmap='gray'); plt.title('Contur (Canny)')
plt.subplot(1, 2, 2); plt.imshow(thresh3, cmap='gray'); plt.title('Regiune identificata')
plt.show()