请用代码实现psnr和ssim的计算

好的，下面给出Python代码实现PSNR和SSIM的计算。

首先，我们需要导入相关的库：numpy、scipy和opencv-python。

import cv2
import numpy as np
from scipy import signal
from scipy.ndimage import gaussian_filter

接着，我们定义计算PSNR的函数psnr()：

def psnr(img1, img2):
    mse = np.mean((img1 - img2) ** 2)
    if mse == 0:
        return 100.0
    PIXEL_MAX = 255.0
    return 20 * np.log10(PIXEL_MAX / np.sqrt(mse))

其中，img1和img2分别为两张图片，函数返回它们的PSNR值。

然后，我们定义计算SSIM的函数ssim()：

def ssim(img1, img2, K=0.01, window_size=11, sigma=1.5):
    img1 = img1.astype(np.float64)
    img2 = img2.astype(np.float64)
    window = _gaussian_window(window_size, sigma)
    K1 = K ** 2
    K2 = K ** 2
    L = 255
    C1 = (K1 * L) ** 2
    C2 = (K2 * L) ** 2
    mu1 = signal.convolve2d(img1, window, mode='valid') / np.sum(window)
    mu2 = signal.convolve2d(img2, window, mode='valid') / np.sum(window)
    mu1_sq = mu1 ** 2
    mu2_sq = mu2 ** 2
    mu1_mu2 = mu1 * mu2
    sigma1_sq = signal.convolve2d(img1 ** 2, window, mode='valid') / np.sum(window) - mu1_sq
    sigma2_sq = signal.convolve2d(img2 ** 2, window, mode='valid') / np.sum(window) - mu2_sq
    sigma12 = signal.convolve2d(img1 * img2, window, mode='valid') / np.sum(window) - mu1_mu2
    ssim_map = ((2 * mu1_mu2 + C1) * (2 * sigma12 + C2)) / ((mu1_sq + mu2_sq + C1) * (sigma1_sq + sigma2_sq + C2))
    return np.mean(ssim_map)

其中，img1和img2分别为两张图片，K、window_size和sigma是SSIM算法中的参数。函数返回它们的SSIM值。

最后，我们定义一个辅助函数_gaussian_window()，用于生成高斯窗口：

def _gaussian_window(window_size, sigma):
    gaussian = np.zeros((window_size, window_size), dtype=np.float64)
    center = window_size // 2
    for i in range(window_size):
        for j in range(window_size):
            gaussian[i, j] = np.exp(-0.5 * ((i - center) ** 2 + (j - center) ** 2) / sigma ** 2)
    return gaussian / np.sum(gaussian)

现在，我们可以用psnr()和ssim()函数来计算两张图片之间的PSNR和SSIM了。

例如，下面的代码演示了如何计算两张图片image1和image2之间的PSNR和SSIM：

image1 = cv2.imread('image1.jpg')
image2 = cv2.imread('image2.jpg')
print('PSNR:', psnr(image1, image2))
print('SSIM:', ssim(image1, image2))