对JPEG压缩后的图像进行滤泼去除压缩导致的块效应，且不能跟去噪滤波一样是全画面滤波，用python写一个程序。

好的，下面是一个基于边界保护和纹理保护的滤波算法的示例代码，用于对JPEG压缩后的图像进行滤泼去除压缩导致的块效应：

import cv2
import numpy as np

# 读取压缩后的JPEG图像
img = cv2.imread('compressed.jpg')

# 分块大小
block_size = 16

# 相邻块之间的重叠区域大小
overlap_size = 4

# 使用基于边界保护的均值滤波算法对图像进行局部滤波
def boundary_mean_filter(block):
    # 计算块的边界
    h, w = block.shape[:2]
    top = block[0, :, :]
    bottom = block[h-1, :, :]
    left = block[:, 0, :]
    right = block[:, w-1, :]

    # 对边界进行均值滤波
    top_mean = np.mean(top, axis=0)
    bottom_mean = np.mean(bottom, axis=0)
    left_mean = np.mean(left, axis=0)
    right_mean = np.mean(right, axis=0)

    # 将边界像素值替换成均值
    block[0:overlap_size, :, :] = top_mean
    block[h-overlap_size:h, :, :] = bottom_mean
    block[:, 0:overlap_size, :] = left_mean
    block[:, w-overlap_size:w, :] = right_mean

    # 对块中心区域进行均值滤波
    center = block[overlap_size:h-overlap_size, overlap_size:w-overlap_size, :]
    center_mean = np.mean(center, axis=(0,1))
    block[overlap_size:h-overlap_size, overlap_size:w-overlap_size, :] = center_mean

    return block

# 使用基于纹理保护的中值滤波算法对图像进行局部滤波
def texture_median_filter(block):
    # 计算块的纹理信息
    dx = cv2.Sobel(block, cv2.CV_64F, 1, 0, ksize=3)
    dy = cv2.Sobel(block, cv2.CV_64F, 0, 1, ksize=3)
    texture = np.abs(dx) + np.abs(dy)

    # 对纹理信息进行中值滤波
    texture_median = np.median(texture)
    texture[texture < texture_median] = 0
    texture[texture >= texture_median] = 1

    # 对块进行中值滤波
    for i in range(3):
        channel = block[:, :, i]
        channel_median = np.median(channel)
        channel[texture == 1] = channel_median

    return block

# 对图像进行分块处理，并对每个块进行局部滤波
h, w = img.shape[:2]
new_img = np.zeros((h, w, 3), dtype=np.uint8)
for i in range(0, h, block_size):
    for j in range(0, w, block_size):
        block = img[i:i+block_size, j:j+block_size, :]
        block = boundary_mean_filter(block)
        block = texture_median_filter(block)
        new_img[i:i+block_size, j:j+block_size, :] = block

# 将处理后的图像保存到文件
cv2.imwrite('filtered.jpg', new_img)

在上述代码中，我们首先读取了压缩后的JPEG图像，并定义了分块和重叠区域的大小。然后，我们定义了两个函数`boundary_mean_filter`和`texture_median_filter`，分别用于基于边界保护和纹理保护的局部滤波。在`boundary_mean_filter`函数中，我们首先计算了块的边界，然后对边界进行均值滤波，最后对块中心区域进行均值滤波。在`texture_median_filter`函数中，我们首先计算了块的纹理信息，然后对纹理信息进行中值滤波，最后对块进行中值滤波。最后，我们对图像进行分块处理，并对每个块进行局部滤波，得到处理后的图像，并将其保存到文件中。

希望这个示例代码能够帮助到您实现对JPEG压缩后的图像进行滤泼去除压缩导致的块效应的功能。