python中Roberts算子是什么

def filter2D(src, ddepth, kernel, dst=None, anchor=None, delta=None, borderType=None)

import cv2 as cv

    import matplotlib.pyplot as plt

    # 读取图像

    img = cv.imread('data.jpg', cv.COLOR_BGR2GRAY)

    rgb_img = cv.cvtColor(img, cv.COLOR_BGR2RGB)

    # 灰度化处理图像

    grayImage = cv.cvtColor(img, cv.COLOR_BGR2GRAY)

    # Roberts 算子

    kernelx = np.array([[-1, 0], [0, 1]], dtype=int)

    kernely = np.array([[0, -1], [1, 0]], dtype=int)

    x = cv.filter2D(grayImage, cv.CV_16S, kernelx)

    y = cv.filter2D(grayImage, cv.CV_16S, kernely)

    # 转 uint8 ,图像融合

    absX = cv.convertScaleAbs(x)

    absY = cv.convertScaleAbs(y)

    Roberts = cv.addWeighted(absX, 0.5, absY, 0.5, 0)

    # 显示图形

    titles = ['原始图像', 'Roberts算子']

    images = [rgb_img, Roberts]

    for i in range(2):

        plt.subplot(1, 2, i + 1), plt.imshow(images[i], 'gray')

        plt.title(titles[i])

        plt.xticks([]), plt.yticks([])

    plt.show()