cv_functions/functions.py

import cv2
import numpy as np
import os

# Load Haar Cascades for face and object detection
face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')
car_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_car.xml')

# Helper function to convert PIL to OpenCV
def pil_to_cv2(image):
    return cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR)

# Helper function to convert OpenCV to PIL
def cv2_to_pil(image):
    return cv2.cvtColor(image, cv2.COLOR_BGR2RGB)

# 1. Image and Video I/O
def image_video_io(image=None, video=None):
    outputs = []
    if image is not None:
        img = pil_to_cv2(image)
        outputs.append(cv2_to_pil(img))
    if video is not None:
        cap = cv2.VideoCapture(video)
        ret, frame = cap.read()
        if ret:
            outputs.append(cv2_to_pil(frame))
        cap.release()
    return outputs

# 2. Color Space Conversion
def color_space_conversion(image, color_space):
    img = pil_to_cv2(image)
    if color_space == "HSV":
        output = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
    elif color_space == "LAB":
        output = cv2.cvtColor(img, cv2.COLOR_BGR2LAB)
    else:
        output = img
    return cv2_to_pil(output)

# 3. Image Resizing and Cropping
def resize_crop(image, scale, crop_x, crop_y, crop_w, crop_h):
    img = pil_to_cv2(image)
    h, w = img.shape[:2]
    new_w, new_h = int(w * scale), int(h * scale)
    resized = cv2.resize(img, (new_w, new_h))
    crop_x, crop_y, crop_w, crop_h = int(crop_x * w), int(crop_y * h), int(crop_w * w), int(crop_h * h)
    cropped = img[crop_y:crop_y+crop_h, crop_x:crop_x+crop_w]
    return [cv2_to_pil(resized), cv2_to_pil(cropped)]

# 4. Geometric Transformations
def geometric_transform(image, angle, tx, ty):
    img = pil_to_cv2(image)
    h, w = img.shape[:2]
    center = (w // 2, h // 2)
    M = cv2.getRotationMatrix2D(center, angle, 1.0)
    M[:, 2] += [tx, ty]
    transformed = cv2.warpAffine(img, M, (w, h))
    return cv2_to_pil(transformed)

# 5. Image Thresholding
def thresholding(image, thresh_type, thresh_value, block_size, C):
    img = pil_to_cv2(image)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    if thresh_type == "Global":
        _, thresh = cv2.threshold(gray, thresh_value, 255, cv2.THRESH_BINARY)
    else:
        thresh = cv2.adaptiveThreshold(gray, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, block_size, C)
    return cv2_to_pil(thresh)

# 6. Edge Detection
def edge_detection(image, edge_type, canny_t1, canny_t2):
    img = pil_to_cv2(image)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    if edge_type == "Canny":
        edges = cv2.Canny(gray, canny_t1, canny_t2)
    elif edge_type == "Sobel":
        sobelx = cv2.Sobel(gray, cv2.CV_64F, 1, 0, ksize=5)
        sobely = cv2.Sobel(gray, cv2.CV_64F, 0, 1, ksize=5)
        edges = cv2.magnitude(sobelx, sobely).astype(np.uint8)
    else:  # Laplacian
        edges = cv2.Laplacian(gray, cv2.CV_64F).astype(np.uint8)
    return cv2_to_pil(edges)

# 7. Image Filtering
def image_filtering(image, filter_type, kernel_size):
    img = pil_to_cv2(image)
    kernel_size = int(kernel_size) | 1
    if filter_type == "Gaussian":
        filtered = cv2.GaussianBlur(img, (kernel_size, kernel_size), 0)
    else:  # Median
        filtered = cv2.medianBlur(img, kernel_size)
    return cv2_to_pil(filtered)

# 8. Contour Detection
def contour_detection(image):
    img = pil_to_cv2(image)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    _, thresh = cv2.threshold(gray, 127, 255, cv2.THRESH_BINARY)
    contours, _ = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
    output = img.copy()
    cv2.drawContours(output, contours, -1, (0, 255, 0), 2)
    return cv2_to_pil(output)

# 9. Feature Detection (ORB)
def feature_detection(image):
    img = pil_to_cv2(image)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    orb = cv2.ORB_create()
    keypoints, _ = orb.detectAndCompute(gray, None)
    output = cv2.drawKeypoints(img, keypoints, None, color=(0, 255, 0), flags=0)
    return cv2_to_pil(output)

# 10. Object Detection (Haar Cascade for cars)
def object_detection(image):
    img = pil_to_cv2(image)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    cars = car_cascade.detectMultiScale(gray, scaleFactor=1.1, minNeighbors=5, minSize=(30, 30))
    output = img.copy()
    for (x, y, w, h) in cars:
        cv2.rectangle(output, (x, y), (x+w, y+h), (0, 255, 0), 2)
    return cv2_to_pil(output)

# 11. Face Detection
def face_detection(image):
    img = pil_to_cv2(image)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    faces = face_cascade.detectMultiScale(gray, scaleFactor=1.1, minNeighbors=5, minSize=(30, 30))
    output = img.copy()
    for (x, y, w, h) in faces:
        cv2.rectangle(output, (x, y), (x+w, y+h), (0, 255, 0), 2)
    return cv2_to_pil(output)

# 12. Image Segmentation (GrabCut)
def image_segmentation(image):
    img = pil_to_cv2(image)
    mask = np.zeros(img.shape[:2], np.uint8)
    bgdModel = np.zeros((1, 65), np.float64)
    fgdModel = np.zeros((1, 65), np.float64)
    rect = (50, 50, img.shape[1]-50, img.shape[0]-50)
    cv2.grabCut(img, mask, rect, bgdModel, fgdModel, 5, cv2.GC_INIT_WITH_RECT)
    mask2 = np.where((mask == 2) | (mask == 0), 0, 1).astype('uint8')
    output = img * mask2[:, :, np.newaxis]
    return cv2_to_pil(output)

# 13. Motion Analysis (Optical Flow)
def optical_flow(video):
    cap = cv2.VideoCapture(video)
    ret, frame1 = cap.read()
    if not ret:
        cap.release()
        return None
    prvs = cv2.cvtColor(frame1, cv2.COLOR_BGR2GRAY)
    hsv = np.zeros_like(frame1)
    hsv[..., 1] = 255
    ret, frame2 = cap.read()
    if not ret:
        cap.release()
        return cv2_to_pil(frame1)
    next = cv2.cvtColor(frame2, cv2.COLOR_BGR2GRAY)
    flow = cv2.calcOpticalFlowFarneback(prvs, next, None, 0.5, 3, 15, 3, 5, 1.2, 0)
    mag, ang = cv2.cartToPolar(flow[..., 0], flow[..., 1])
    hsv[..., 0] = ang * 180 / np.pi / 2
    hsv[..., 2] = cv2.normalize(mag, None, 0, 255, cv2.NORM_MINMAX)
    output = cv2.cvtColor(hsv, cv2.COLOR_HSV2BGR)
    cap.release()
    return cv2_to_pil(output)

# 14. Camera Calibration (Simplified)
def camera_calibration(image):
    img = pil_to_cv2(image)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    chessboard_size = (9, 6)
    ret, corners = cv2.findChessboardCorners(gray, chessboard_size, None)
    output = img.copy()
    if ret:
        cv2.drawChessboardCorners(output, chessboard_size, corners, ret)
    return cv2_to_pil(output)

# 15. Stereo Vision (Simplified Disparity Map)
def stereo_vision(image):
    img = pil_to_cv2(image)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    stereo = cv2.StereoBM_create(numDisparities=16, blockSize=15)
    disparity = stereo.compute(gray, gray)  # Simplified: using same image
    disparity = cv2.normalize(disparity, None, 0, 255, cv2.NORM_MINMAX).astype(np.uint8)
    return cv2_to_pil(disparity)

# 16. Background Subtraction
def background_subtraction(video):
    cap = cv2.VideoCapture(video)
    fgbg = cv2.createBackgroundSubtractorMOG2()
    ret, frame = cap.read()
    if not ret:
        cap.release()
        return None
    fgmask = fgbg.apply(frame)
    output = cv2.cvtColor(fgmask, cv2.COLOR_GRAY2BGR)
    cap.release()
    return cv2_to_pil(output)

# 17. Image Stitching
def image_stitching(image1, image2):
    img1 = pil_to_cv2(image1)
    img2 = pil_to_cv2(image2)
    gray1 = cv2.cvtColor(img1, cv2.COLOR_BGR2GRAY)
    gray2 = cv2.cvtColor(img2, cv2.COLOR_BGR2GRAY)
    orb = cv2.ORB_create()
    kp1, des1 = orb.detectAndCompute(gray1, None)
    kp2, des2 = orb.detectAndCompute(gray2, None)
    bf = cv2.BFMatcher(cv2.NORM_HAMMING, crossCheck=True)
    matches = bf.match(des1, des2)
    matches = sorted(matches, key=lambda x: x.distance)
    good_matches = matches[:10]
    src_pts = np.float32([kp1[m.queryIdx].pt for m in good_matches]).reshape(-1, 1, 2)
    dst_pts = np.float32([kp2[m.trainIdx].pt for m in good_matches]).reshape(-1, 1, 2)
    M, _ = cv2.findHomography(src_pts, dst_pts, cv2.RANSAC, 5.0)
    h, w = img1.shape[:2]
    result = cv2.warpPerspective(img2, M, (w * 2, h))
    result[0:h, 0:w] = img1
    return cv2_to_pil(result)

# 18. Machine Learning (K-Means)
def kmeans_clustering(image, k):
    img = pil_to_cv2(image)
    Z = img.reshape((-1, 3))
    Z = np.float32(Z)
    criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 10, 1.0)
    _, label, center = cv2.kmeans(Z, k, None, criteria, 10, cv2.KMEANS_RANDOM_CENTERS)
    center = np.uint8(center)
    res = center[label.flatten()]
    output = res.reshape(img.shape)
    return cv2_to_pil(output)

# 19. Deep Learning (MobileNet SSD)
def deep_learning(image, prototxt_file, model_file):
    if prototxt_file is None or model_file is None:
        return None  # Model files must be uploaded
    img = pil_to_cv2(image)
    net = cv2.dnn.readNetFromCaffe(prototxt_file.file.name, model_file.file.name)
    blob = cv2.dnn.blobFromImage(cv2.resize(img, (300, 300)), 0.007843, (300, 300), 127.5)
    net.setInput(blob)
    detections = net.forward()
    output = img.copy()
    for i in range(detections.shape[2]):
        confidence = detections[0, 0, i, 2]
        if confidence > 0.5:
            box = detections[0, 0, i, 3:7] * np.array([img.shape[1], img.shape[0], img.shape[1], img.shape[0]])
            (startX, startY, endX, endY) = box.astype("int")
            cv2.rectangle(output, (startX, startY), (endX, endY), (0, 255, 0), 2)
    return cv2_to_pil(output)

# 20. Drawing and Text
def drawing_text(image, shape, text):
    img = pil_to_cv2(image)
    output = img.copy()
    h, w = img.shape[:2]
    if shape == "Rectangle":
        cv2.rectangle(output, (50, 50), (w-50, h-50), (0, 255, 0), 2)
    elif shape == "Circle":
        cv2.circle(output, (w//2, h//2), 50, (0, 255, 0), 2)
    cv2.putText(output, text, (50, 50), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 0, 0), 2)
    return cv2_to_pil(output)