Python opencv图像处理基础总结(五) 图像金字塔 图像梯度 Canny算法边缘提取

import cv2 as cv


# 高斯金字塔
def pyramid_demo(image):
    level = 3  # 金字塔层数
    temp = image.copy()
    pyramid_images = []
    for i in range(level):
        dst = cv.pyrDown(temp)       # 降采样
        pyramid_images.append(dst)   # 降采样的结果添加进列表
        cv.imshow('pyramid_down' + str(i), dst)   # 金字塔第几层  imshow
        temp = dst.copy()            # 采样的图像又赋给temp  接着降采样
    return pyramid_images


# 拉普拉斯金字塔
# 由高斯金字塔可以构建拉普拉斯金字塔
def lapalian_demo(image):
    pyramid_images = pyramid_demo(image)
    level = len(pyramid_images)         # 求层数
    for i in range(level - 1, -1, -1):  # 每次递减
        if (i - 1) < 0:
            expand = cv.pyrUp(pyramid_images[i], dstsize=image.shape[:2])  # 升采样
            lpls = cv.subtract(image, expand)
            cv.imshow("lapalian_down" + str(i), lpls)
        else:
            expand = cv.pyrUp(pyramid_images[i], dstsize=pyramid_images[i - 1].shape[:2])
            lpls = cv.subtract(pyramid_images[i - 1], expand)
            cv.imshow("lapalian_down_" + str(i), lpls)


if __name__ == '__main__':
    img = cv.imread(r'./test/004.jpg')  # 图片须是2^n 格式大小  例如（512*512）的
    cv.imshow("input image", img)
    lapalian_demo(img)
    cv.waitKey(0)
    cv.destroyAllWindows()

cv2.Sobel(src, ddepth, dx, dy, dst=None, ksize=None, scale=None, delta=None, borderType=None)

import cv2


def sobel_demo(image):
    grad_x = cv2.Sobel(image, cv2.CV_32F, 1, 0)   # x方向一阶导数
    grad_y = cv2.Sobel(image, cv2.CV_32F, 0, 1)   # y方向一阶导数
    # 一阶导数算出来可能有正负  最后全部转到8位的图像上去
    gradx = cv2.convertScaleAbs(grad_x)
    grady = cv2.convertScaleAbs(grad_y)
    cv2.imshow("gradient_x", gradx)
    cv2.imshow("gradient_y", grady)
    # 计算两个数组的加权和
    gradxy = cv2.addWeighted(gradx, 0.5, grady, 0.5, 0)
    cv2.imshow("gradient_xy", gradxy)


if __name__ == "__main__":
    src = cv2.imread(r"./test/018.jpg")
    src = cv2.resize(src, None, fx=0.5, fy=0.5)
    cv2.imshow("image", src)
    sobel_demo(src)
    cv2.waitKey(0)
    cv2.destroyAllWindows()

import cv2


def scharr_demo(image):
    grad_x = cv2.Scharr(image, cv2.CV_32F, 1, 0)   # x方向一阶导数
    grad_y = cv2.Scharr(image, cv2.CV_32F, 0, 1)   # y方向一阶导数
    # 一阶导数算出来可能有正负  最后全部转到8位的图像上去
    gradx = cv2.convertScaleAbs(grad_x)
    grady = cv2.convertScaleAbs(grad_y)
    cv2.imshow("gradient_x", gradx)
    cv2.imshow("gradient_y", grady)
    # 计算两个数组的加权和
    gradxy = cv2.addWeighted(gradx, 0.5, grady, 0.5, 0)
    cv2.imshow("gradient_xy", gradxy)


if __name__ == "__main__":
    src = cv2.imread(r"./test/018.jpg")
    src = cv2.resize(src, None, fx=0.5, fy=0.5)
    cv2.imshow("image", src)
    scharr_demo(src)
    cv2.waitKey(0)
    cv2.destroyAllWindows()

cv2.Laplacian(src, ddepth, dst=None, ksize=None, scale=None, delta=None, borderType=None)

import cv2 as cv
import numpy as np


# 拉普拉斯算子
def Laplace_demo(image):
    dst = cv.Laplacian(image, cv.CV_32F)
    lpls_1 = cv.convertScaleAbs(dst)
    cv.imshow("Laplace_1", lpls_1)
    # 自定义拉普拉斯算子
    kernel = np.array([[1, 1, 1], [1, -8, 1], [1, 1, 1]])
    dst = cv.filter2D(image, cv.CV_32F, kernel)
    lpls_2 = cv.convertScaleAbs(dst)
    cv.imshow("Laplace_2", lpls_2)


if __name__ == "__main__":
    src = cv.imread(r"./test/018.jpg")
    src = cv.resize(src, None, fx=0.5, fy=0.5)
    cv.imshow("image", src)
    Laplace_demo(src)
    cv.waitKey(0)
    cv.destroyAllWindows()

cv2.Canny(image, threshold1, threshold2, edges=None, apertureSize=None, L2gradient=None):

import cv2 as cv


def edge_demo(image):
    # 高斯模糊  降低噪声
    blurred = cv.GaussianBlur(image, (3, 3), 0)
    # 转为灰度图像
    gray = cv.cvtColor(blurred, cv.COLOR_BGR2GRAY)
    # 计算x y 方向梯度
    grad_x = cv.Sobel(gray, cv.CV_16SC1, 1, 0)
    grad_y = cv.Sobel(gray, cv.CV_16SC1, 0, 1)
    edge_output = cv.Canny(grad_x, grad_y,  50, 100)
    cv.imshow('edge image_1', edge_output)
    # 彩色图像
    dist = cv.bitwise_and(image, image, mask=edge_output)
    cv.imshow('edge image_2', dist)


if __name__ == "__main__":
    src = cv.imread(r"./test/033.png")
    src = cv.resize(src, None, fx=0.5, fy=0.5)
    cv.imshow("image", src)
    edge_demo(src)
    cv.waitKey(0)
    cv.destroyAllWindows()