基于OpenCV的图像边缘检测与轮廓分析

import cv2

pic = cv2.imread(r"picture_1.jpg", 0)
pic_x_64 = cv2.Sobel(pic, cv2.CV_64F, dx=1, dy=0)
pic_x_full = cv2.convertScaleAbs(pic_x_64)
pic_y_64 = cv2.Sobel(pic, cv2.CV_64F, dx=0, dy=1)
pic_y_full = cv2.convertScaleAbs(pic_y_64)
pic_xy_sobel_full = cv2.addWeighted(pic_x_full, 1, pic_y_full, 1, 0)

pic = cv2.imread(r"picture_1.jpg", cv2.IMREAD_GRAYSCALE)
pic_x_64 = cv2.Scharr(pic, cv2.CV_64F, dx=1, dy=0)
pic_x_full = cv2.convertScaleAbs(pic_x_64)
pic_y_64 = cv2.Scharr(pic, cv2.CV_64F, dx=0, dy=1)
pic_y_full = cv2.convertScaleAbs(pic_y_64)
pic_xy_scharr_full = cv2.addWeighted(pic_x_full, 1, pic_y_full, 1, 0)

pic = cv2.imread(r"picture_1.jpg", cv2.IMREAD_GRAYSCALE)
pic_lap = cv2.Laplacian(pic, cv2.CV_64F, ksize=3)
pic_lap_full = cv2.convertScaleAbs(pic_lap)

pic = cv2.imread(r"picture_1.jpg", cv2.IMREAD_GRAYSCALE)
pic_canny = cv2.Canny(pic, 100, 150)

# 灰度图处理
phone = cv2.imread(r"picture_1.jpg")
phone = cv2.resize(phone, dsize=None, fx=0.4, fy=0.4)
phone_gray = cv2.cvtColor(phone, cv2.COLOR_BGR2GRAY)

# 阈值处理为二值图像
ret, phone_binary = cv2.threshold(phone_gray, 120, 255, cv2.THRESH_BINARY)

# 查找轮廓
_, contours, hierarchy = cv2.findContours(phone_binary, cv2.RETR_TREE, cv2.CHAIN_APPROX_NONE)

print(len(contours))

image_copy = phone.copy()
cv2.drawContours(image=image_copy, contours=contours, contourIdx=-1, color=(0, 255, 0), thickness=2)

# 逐个绘制轮廓的替代方法
for i in range(len(contours)):
    image_copy = cv2.drawContours(image=image_copy, contours=contours, contourIdx=i, color=(0, 255, 0), thickness=3)
    cv2.imshow('Contours_show', image_copy)
    cv2.waitKey(0)

# 轮廓面积计算
area_0 = cv2.contourArea(contours[0])
area_1 = cv2.contourArea(contours[1])

# 轮廓周长计算
length = cv2.arcLength(contours[0], closed=True)

# 根据面积筛选轮廓
a_list = []
for i in contours:
    if cv2.contourArea(i) > 10000:
        a_list.append(i)

# 根据面积排序获取最大轮廓
sortcnt = sorted(contours, key=cv2.contourArea, reverse=True)[0]
image_contours = cv2.drawContours(image_copy, [sortcnt], contourIdx=-1, color=(0, 0, 255), thickness=3)

cnt = contours[8]

# 最小外接圆
(x, y), r = cv2.minEnclosingCircle(cnt)
phone_circle = cv2.circle(phone, center=(int(x), int(y)), radius=int(r), color=(0, 255, 0), thickness=2)

# 最小外接矩形
x, y, w, h = cv2.boundingRect(cnt)
phone_rectangle = cv2.rectangle(phone, pt1=(x, y), pt2=(x + w, y + h), color=(0, 255, 0), thickness=2)

phone = cv2.imread('phone.png')
phone_gray = cv2.cvtColor(phone, cv2.COLOR_BGR2GRAY)
ret, phone_thresh = cv2.threshold(phone_gray, 120, 255, cv2.THRESH_BINARY)

# 兼容不同OpenCV版本的轮廓查找
contours = cv2.findContours(phone_thresh, cv2.RETR_TREE, cv2.CHAIN_APPROX_NONE)[-2]

# 轮廓近似
epsilon = 0.01 * cv2.arcLength(contours[0], closed=True)
approx = cv2.approxPolyDP(contours[0], epsilon, closed=True)

print(contours[0].shape)  # 原始轮廓点数
print(approx.shape)       # 近似后轮廓点数

# 绘制近似轮廓
phone_new = phone.copy()
image_contours = cv2.drawContours(phone_new, [approx], contourIdx=-1, color=(0, 255, 0), thickness=3)