Harris角点检测python实现

Harris角点检测是一种常用的图像处理算法，用于检测图像中的角点。以下是用Python实现Harris角点检测的代码：

import cv2
import numpy as np

def Harris_corner_detection(image, k=0.04, threshold=0.01):
    # Convert the image to grayscale
    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    height, width = gray.shape

    # Calculate the gradients in x and y directions
    Ix = cv2.Sobel(gray, cv2.CV_64F, 1, 0, ksize=3)
    Iy = cv2.Sobel(gray, cv2.CV_64F, 0, 1, ksize=3)

    # Calculate the products of the gradients
    Ix2 = np.multiply(Ix, Ix)
    Iy2 = np.multiply(Iy, Iy)
    Ixy = np.multiply(Ix, Iy)

    # Calculate the sums of the products of gradients
    Sx2 = cv2.GaussianBlur(Ix2, (3, 3), 0)
    Sy2 = cv2.GaussianBlur(Iy2, (3, 3), 0)
    Sxy = cv2.GaussianBlur(Ixy, (3, 3), 0)

    # Calculate the Harris response for each pixel
    R = np.zeros((height, width))
    for y in range(height):
        for x in range(width):
            det = (Sx2[y, x] * Sy2[y, x]) - (Sxy[y, x] ** 2)
            trace = Sx2[y, x] + Sy2[y, x]
            R[y, x] = det - (k * (trace ** 2))

    # Threshold the response image and mark the corners
    image_copy = image.copy()
    for y in range(height):
        for x in range(width):
            if R[y, x] > threshold * R.max():
                cv2.circle(image_copy, (x, y), 3, (0, 0, 255), -1)

    return image_copy

# Test the function on an image
image = cv2.imread('test.jpg')
corners = Harris_corner_detection(image)
cv2.imshow('Harris Corners', corners)
cv2.waitKey(0)
cv2.destroyAllWindows()

在上述代码中，我们首先将输入图像转换为灰度图像。然后，我们使用cv2.Sobel()函数计算图像在x和y方向上的梯度。接下来，我们计算梯度的乘积并使用cv2.GaussianBlur()函数对它们进行平滑处理。最后，我们计算每个像素的Harris响应，并在响应图像中标记角点。最后，我们在测试图像上调用该函数并显示结果。

需要注意的是，Harris角点检测的参数k和阈值threshold可以根据具体情况调整。