OpenCV-Python之图像金字塔

版权声明：本文为博主原创文章，转载请标明出处，蟹蟹 https://blog.csdn.net/fly_wt/article/details/85007972

图像金字塔化：先进行图像平滑，再进行降采样，根据降采样率，得到一系列尺寸逐渐减小的图像。
操作：操作：n次（高斯卷积->2倍降采样）->n层金字塔
目的：捕捉不同尺寸的物体
直接降采样会损失信息，所以在降采样之前要进行高斯滤波

def pyramid_demo(img):
    level = 3
    temp = img.copy()
    pyramid_image = []
    for i in range(level):
        dst = cv.pyrDown(temp)
        pyramid_image.append(dst)
        cv.imshow('pyramid_down_'+str(i), dst)
        temp = dst.copy()
    return pyramid_image

src = cv.imread('./data/lena.jpg',1)
cv.imshow('source image', src)
pyramid_demo(src)
cv.waitKey(0)
cv.destroyAllWindows()

结果如下：

Laplace金字塔

def Laplace_demo(image):
    pyramid_image = pyramid_demo(image)
    level = len(pyramid_image)
    for i in range(level-1, -1, -1):
        if (i-1) < 0:
            expand = cv.pyrUp(pyramid_image[i], dstsize=image.shape[:2])
            laplace = cv.subtract(image, expand)
            cv.imshow('Laplace_Image_'+str(i), laplace)
        else:
            expand = cv.pyrUp(pyramid_image[i], dstsize=pyramid_image[i - 1].shape[:2])
            laplace = cv.subtract(pyramid_image[i - 1], expand)
            cv.imshow('Laplace_Image_' + str(i), laplace)