请帮我写一个opencv python 程序，计算图片经过均值滤波、中值滤波、自适应中值滤波、两次小波分解滤波后的MSE,P SNR值，从而评价这几个滤波处理的结果的好坏

在OpenCV中，计算图像处理后（如滤波）的MSE（均方误差）和PSNR（峰值信噪比）通常用于比较原始图像和处理后的图像质量。以下是使用Python和OpenCV来实现这个功能的一个简单示例，涵盖了你提到的几种滤波方法：

import cv2
import numpy as np
from skimage.measure import compare_ssim
from scipy.fftpack import dct, idct

def calculate_mse(original_image, processed_image):
    mse = np.mean((original_image - processed_image) ** 2)
    return mse

def calculate_psnr(original_image, processed_image):
    mse = calculate_mse(original_image, processed_image)
    max_pixel_value = 255.0  # 对于8位灰度图像
    psnr = 10 * np.log10(max_pixel_value**2 / mse)
    return psnr

# 假设你已经有了原始图像和滤波后的图像
original_image = cv2.imread('original.jpg', 0)  # 读取灰度图像
processed_images = {
    "mean": cv2.blur(original_image, (3, 3)),  # 均值滤波
    "median": cv2.medianBlur(original_image, 5),  # 中值滤波
    "adaptive_median": cv2.adaptiveMedianFilter(original_image, (5, 5), cv2.ADAPTIVE_THRESH_GAUSSIAN_C),  # 自适应中值滤波
    "wavelet": np.abs(dct(dct(original_image, norm="ortho", axis=-1), norm="ortho", axis=-2)),  # 二次小波分解滤波
}

mse_values = []
psnr_values = []

for name, filtered_image in processed_images.items():
    mse = calculate_mse(original_image, filtered_image)
    psnr = calculate_psnr(original_image, filtered_image)
    mse_values.append(mse)
    psnr_values.append(psnr)
    print(f"{name} MSE: {mse:.4f}, PSNR: {psnr:.4f}")

# 结果总结
print("\nOverall comparison:")
for i, (method, (mse, psnr)) in enumerate(zip(processed_images.keys(), zip(mse_values, psnr_values))):
    print(f"{i+1}. {method}: MSE = {mse:.4f}, PSNR = {psnr:.4f}")
    
# 如果需要，可以使用skimage.metrics.structural_similarity()计算SSIM作为额外的评估指标
ssim_values = [compare_ssim(original_image, filtered_image, multichannel=True) for filtered_image in processed_images.values()]
print("SSIM values:", ssim_values)

这个程序首先定义了计算MSE和PSNR的方法，然后读取原始图像并应用各种滤波操作。最后，它计算每种滤波方法下的MSE和PSNR值，并输出对比结果。