OpenCV 介绍和使用

OpenCV (Open Source Computer Vision Library) 是一个开源的计算机视觉和机器学习软件库。本教程将带你从基础到高级全面学习 OpenCV 的 Python 接口。

目录

1. 安装与环境配置
2. 图像基础操作
3. 图像处理技术
4. 特征检测与描述
5. 视频处理
6. 对象检测
7. 深度学习与OpenCV
8. 实战项目

安装与环境配置

安装OpenCV

# 安装OpenCV主模块
pip install opencv-python

# 安装包含额外模块的版本（如SIFT等）
pip install opencv-contrib-python

验证安装

import cv2

print(cv2.__version__)  # 打印OpenCV版本

图像基础操作

读取和显示图像

import cv2

# 读取图像
img = cv2.imread('image.jpg')  # 参数：文件路径, 标志(cv2.IMREAD_COLOR/GRAYSCALE/UNCHANGED)

# 显示图像
cv2.imshow('Image Window', img)
cv2.waitKey(0)  # 等待任意按键
cv2.destroyAllWindows()  # 关闭所有窗口

# 保存图像
cv2.imwrite('output.jpg', img)

图像属性

print("图像形状:", img.shape)  # (高度, 宽度, 通道数)
print("图像大小:", img.size)   # 总像素数
print("数据类型:", img.dtype)  # 数据类型

颜色空间转换

# BGR转灰度图
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

# BGR转HSV
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)

# 显示转换结果
cv2.imshow('Gray Image', gray)
cv2.imshow('HSV Image', hsv)
cv2.waitKey(0)

图像缩放与裁剪

# 缩放
resized = cv2.resize(img, (new_width, new_height))  # 指定新尺寸
resized = cv2.resize(img, None, fx=0.5, fy=0.5)    # 按比例缩放

# 裁剪
cropped = img[y1:y2, x1:x2]  # y1,y2是高度范围，x1,x2是宽度范围

图像处理技术

图像阈值处理

# 简单阈值
ret, thresh1 = cv2.threshold(gray, 127, 255, cv2.THRESH_BINARY)

# 自适应阈值
thresh2 = cv2.adaptiveThreshold(gray, 255, cv2.ADAPTIVE_THRESH_MEAN_C, 
                               cv2.THRESH_BINARY, 11, 2)

# Otsu's二值化
ret, thresh3 = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)

图像滤波

# 高斯模糊
blur = cv2.GaussianBlur(img, (5,5), 0)

# 中值滤波
median = cv2.medianBlur(img, 5)

# 双边滤波
bilateral = cv2.bilateralFilter(img, 9, 75, 75)

边缘检测

# Canny边缘检测
edges = cv2.Canny(img, 100, 200)  # 参数：低阈值，高阈值

# Sobel算子
sobelx = cv2.Sobel(gray, cv2.CV_64F, 1, 0, ksize=5)
sobely = cv2.Sobel(gray, cv2.CV_64F, 0, 1, ksize=5)

形态学操作

kernel = np.ones((5,5), np.uint8)

# 腐蚀
erosion = cv2.erode(img, kernel, iterations=1)

# 膨胀
dilation = cv2.dilate(img, kernel, iterations=1)

# 开运算（先腐蚀后膨胀）
opening = cv2.morphologyEx(img, cv2.MORPH_OPEN, kernel)

# 闭运算（先膨胀后腐蚀）
closing = cv2.morphologyEx(img, cv2.MORPH_CLOSE, kernel)

特征检测与描述

Harris角点检测

gray = np.float32(gray)
dst = cv2.cornerHarris(gray, 2, 3, 0.04)
dst = cv2.dilate(dst, None)
img[dst > 0.01 * dst.max()] = [0,0,255]  # 标记角点

SIFT特征检测

sift = cv2.SIFT_create()
keypoints, descriptors = sift.detectAndCompute(gray, None)

# 绘制关键点
img_sift = cv2.drawKeypoints(img, keypoints, None)
cv2.imshow('SIFT Features', img_sift)
cv2.waitKey(0)

SURF特征检测

surf = cv2.xfeatures2d.SURF_create(400)
keypoints, descriptors = surf.detectAndCompute(img, None)

# 绘制关键点
img_surf = cv2.drawKeypoints(img, keypoints, None, (255,0,0), 4)
cv2.imshow('SURF Features', img_surf)
cv2.waitKey(0)

ORB特征检测

orb = cv2.ORB_create()
keypoints, descriptors = orb.detectAndCompute(img, None)

# 绘制关键点
img_orb = cv2.drawKeypoints(img, keypoints, None, color=(0,255,0), flags=0)
cv2.imshow('ORB Features', img_orb)
cv2.waitKey(0)

特征匹配

# 创建匹配器
bf = cv2.BFMatcher(cv2.NORM_HAMMING, crossCheck=True)

# 匹配描述子
matches = bf.match(des1, des2)

# 按距离排序
matches = sorted(matches, key=lambda x:x.distance)

# 绘制匹配结果
img_matches = cv2.drawMatches(img1, kp1, img2, kp2, matches[:10], None, flags=2)
cv2.imshow('Feature Matches', img_matches)
cv2.waitKey(0)

视频处理

读取和显示视频

cap = cv2.VideoCapture('video.mp4')  # 也可以传入0表示摄像头

while cap.isOpened():
    ret, frame = cap.read()
    if not ret:
        break

    # 处理帧
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

    cv2.imshow('Video', gray)

    if cv2.waitKey(25) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()

保存视频

fourcc = cv2.VideoWriter_fourcc(*'XVID')
out = cv2.VideoWriter('output.avi', fourcc, 20.0, (640,480))

while cap.isOpened():
    ret, frame = cap.read()
    if not ret:
        break

    # 处理帧
    out.write(frame)

    cv2.imshow('frame', frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
out.release()
cv2.destroyAllWindows()

背景减除

# 创建背景减除器
fgbg = cv2.createBackgroundSubtractorMOG2()

while True:
    ret, frame = cap.read()
    if not ret:
        break

    # 应用背景减除
    fgmask = fgbg.apply(frame)

    cv2.imshow('Original', frame)
    cv2.imshow('Foreground', fgmask)

    if cv2.waitKey(30) & 0xFF == ord('q'):
        break

对象检测

人脸检测

# 加载预训练的人脸检测模型
face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')

# 检测人脸
faces = face_cascade.detectMultiScale(gray, scaleFactor=1.1, minNeighbors=5)

# 绘制矩形框
for (x,y,w,h) in faces:
    cv2.rectangle(img, (x,y), (x+w,y+h), (255,0,0), 2)

cv2.imshow('Face Detection', img)
cv2.waitKey(0)

行人检测

# 加载HOG描述子和SVM分类器
hog = cv2.HOGDescriptor()
hog.setSVMDetector(cv2.HOGDescriptor_getDefaultPeopleDetector())

# 检测行人
boxes, weights = hog.detectMultiScale(img, winStride=(4,4), padding=(8,8), scale=1.05)

# 绘制检测结果
for (x,y,w,h) in boxes:
    cv2.rectangle(img, (x,y), (x+w,y+h), (0,255,0), 2)

cv2.imshow('Pedestrian Detection', img)
cv2.waitKey(0)

使用深度学习模型进行对象检测

# 加载预训练的MobileNet SSD模型
net = cv2.dnn.readNetFromCaffe('deploy.prototxt', 'mobilenet_iter_73000.caffemodel')
classes = ["background", "aeroplane", "bicycle", "bird", "boat",
           "bottle", "bus", "car", "cat", "chair", "cow", "diningtable",
           "dog", "horse", "motorbike", "person", "pottedplant", "sheep",
           "sofa", "train", "tvmonitor"]

# 准备输入图像
blob = cv2.dnn.blobFromImage(cv2.resize(img, (300,300)), 0.007843, (300,300), 127.5)

# 进行预测
net.setInput(blob)
detections = net.forward()

# 解析检测结果
for i in range(detections.shape[2]):
    confidence = detections[0, 0, i, 2]
    if confidence > 0.5:  # 置信度阈值
        idx = int(detections[0, 0, i, 1])
        box = detections[0, 0, i, 3:7] * np.array([w, h, w, h])
        (startX, startY, endX, endY) = box.astype("int")

        # 绘制边界框和标签
        label = "{}: {:.2f}%".format(classes[idx], confidence * 100)
        cv2.rectangle(img, (startX, startY), (endX, endY), (0,255,0), 2)
        cv2.putText(img, label, (startX, startY-10), 
                   cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0,255,0), 2)

cv2.imshow("Object Detection", img)
cv2.waitKey(0)

深度学习与OpenCV

加载和使用预训练模型

# 加载TensorFlow模型
net = cv2.dnn.readNetFromTensorflow('frozen_inference_graph.pb', 'graph.pbtxt')

# 加载PyTorch模型
net = cv2.dnn.readNetFromTorch('model.pt')

# 使用模型进行预测
blob = cv2.dnn.blobFromImage(img, scalefactor=1.0, size=(224,224), 
                            mean=(104,117,123), swapRB=False, crop=False)
net.setInput(blob)
output = net.forward()

实时风格迁移

# 加载风格迁移模型
net = cv2.dnn.readNetFromTorch('models/instance_norm/mosaic.t7')

while True:
    ret, frame = cap.read()
    if not ret:
        break

    # 准备输入
    blob = cv2.dnn.blobFromImage(frame, 1.0, (frame.shape[1], frame.shape[0]),
                               (103.939, 116.779, 123.680), swapRB=False, crop=False)

    # 应用风格迁移
    net.setInput(blob)
    output = net.forward()

    # 后处理
    output = output.reshape((3, output.shape[2], output.shape[3]))
    output[0] += 103.939
    output[1] += 116.779
    output[2] += 123.680
    output = output.transpose(1, 2, 0)
    output = np.clip(output, 0, 255).astype('uint8')

    cv2.imshow('Styled', output)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

实战项目

1. 实时人脸识别

# 加载人脸检测和识别模型
face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')
recognizer = cv2.face.LBPHFaceRecognizer_create()
recognizer.read('trainer.yml')  # 预训练的人脸识别模型

while True:
    ret, frame = cap.read()
    if not ret:
        break

    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    faces = face_cascade.detectMultiScale(gray, scaleFactor=1.1, minNeighbors=5)

    for (x,y,w,h) in faces:
        roi_gray = gray[y:y+h, x:x+w]

        # 进行识别
        id_, confidence = recognizer.predict(roi_gray)

        # 绘制结果
        cv2.rectangle(frame, (x,y), (x+w,y+h), (255,0,0), 2)
        cv2.putText(frame, f"ID: {id_} Conf: {confidence:.2f}", 
                   (x,y-10), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (255,0,0), 2)

    cv2.imshow('Face Recognition', frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

2. 文档扫描仪

def order_points(pts):
    rect = np.zeros((4, 2), dtype="float32")
    s = pts.sum(axis=1)
    rect[0] = pts[np.argmin(s)]
    rect[2] = pts[np.argmax(s)]

    diff = np.diff(pts, axis=1)
    rect[1] = pts[np.argmin(diff)]
    rect[3] = pts[np.argmax(diff)]
    return rect

def four_point_transform(image, pts):
    rect = order_points(pts)
    (tl, tr, br, bl) = rect

    widthA = np.sqrt(((br[0] - bl[0]) ** 2) + ((br[1] - bl[1]) ** 2))
    widthB = np.sqrt(((tr[0] - tl[0]) ** 2) + ((tr[1] - tl[1]) ** 2))
    maxWidth = max(int(widthA), int(widthB))

    heightA = np.sqrt(((tr[0] - br[0]) ** 2) + ((tr[1] - br[1]) ** 2))
    heightB = np.sqrt(((tl[0] - bl[0]) ** 2) + ((tl[1] - bl[1]) ** 2))
    maxHeight = max(int(heightA), int(heightB))

    dst = np.array([
        [0, 0],
        [maxWidth - 1, 0],
        [maxWidth - 1, maxHeight - 1],
        [0, maxHeight - 1]], dtype="float32")

    M = cv2.getPerspectiveTransform(rect, dst)
    warped = cv2.warpPerspective(image, M, (maxWidth, maxHeight))
    return warped

# 读取图像
image = cv2.imread('document.jpg')
ratio = image.shape[0] / 500.0
orig = image.copy()
image = cv2.resize(image, (int(image.shape[1]/ratio), 500))

# 预处理
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
gray = cv2.GaussianBlur(gray, (5,5), 0)
edged = cv2.Canny(gray, 75, 200)

# 寻找轮廓
cnts = cv2.findContours(edged.copy(), cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)
cnts = cnts[0] if len(cnts) == 2 else cnts[1]
cnts = sorted(cnts, key=cv2.contourArea, reverse=True)[:5]

# 寻找文档轮廓
for c in cnts:
    peri = cv2.arcLength(c, True)
    approx = cv2.approxPolyDP(c, 0.02 * peri, True)

    if len(approx) == 4:
        screenCnt = approx
        break

# 应用透视变换
warped = four_point_transform(orig, screenCnt.reshape(4,2) * ratio)

# 二值化
warped = cv2.cvtColor(warped, cv2.COLOR_BGR2GRAY)
warped = cv2.adaptiveThreshold(warped, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,
                              cv2.THRESH_BINARY, 11, 2)

cv2.imshow("Original", cv2.resize(orig, (500, int(orig.shape[0]/ratio))))
cv2.imshow("Scanned", warped)
cv2.waitKey(0)

3. 车牌识别

# 加载预训练的车牌检测模型
plate_cascade = cv2.CascadeClassifier('haarcascade_russian_plate_number.xml')

while True:
    ret, frame = cap.read()
    if not ret:
        break

    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    plates = plate_cascade.detectMultiScale(gray, 1.1, 4)

    for (x,y,w,h) in plates:
        # 提取车牌区域
        plate = frame[y:y+h, x:x+w]

        # 预处理车牌图像
        plate_gray = cv2.cvtColor(plate, cv2.COLOR_BGR2GRAY)
        _, plate_thresh = cv2.threshold(plate_gray, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)

        # 使用Tesseract OCR识别文本
        text = pytesseract.image_to_string(plate_thresh, config='--psm 7')

        # 绘制结果
        cv2.rectangle(frame, (x,y), (x+w,y+h), (0,255,0), 2)
        cv2.putText(frame, text.strip(), (x,y-10), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (0,255,0), 2)

    cv2.imshow('License Plate Recognition', frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

总结

本教程涵盖了OpenCV Python接口的主要功能，包括：

1. 图像基础操作（读取、显示、保存）
2. 图像处理技术（滤波、边缘检测、形态学操作）
3. 特征检测与匹配（SIFT、SURF、ORB）
4. 视频处理（读取、保存、背景减除）
5. 对象检测（人脸、行人、车牌等）
6. 深度学习与OpenCV结合
7. 实战项目（人脸识别、文档扫描、车牌识别）

要深入学习OpenCV，建议：

1. 实践更多项目
2. 阅读OpenCV官方文档
3. 探索OpenCV的contrib模块
4. 学习计算机视觉基础理论
5. 关注OpenCV的更新和新功能

希望本教程能帮助你掌握OpenCV Python编程！