图像滤波方法

1.图像滤波

代码如下（示例）：

import imageio 
import numpy as np
import torch
import cv2 as cv
from PIL import Image
##图像resize
size=(800,800)
# with Image.open('./data/r_1.png') as image:
#     resized_image = image.resize(size)
# resized_image.save('./data/r_2.png')


# 低通滤波
def Low_Pass_Filter(srcImg_path):
    #img = cv.imread('srcImg_path', 0)
    img = np.array(Image.open(srcImg_path))
    img = cv.cvtColor(img,cv.COLOR_BGR2GRAY)

    # 傅里叶变换
    dft = cv.dft(np.float32(img), flags = cv.DFT_COMPLEX_OUTPUT)
    fshift = np.fft.fftshift(dft)

    # 设置低通滤波器
    rows, cols = img.shape
    crow, ccol = int(rows/2), int(cols/2) # 中心位置
    mask = np.zeros((rows, cols, 2), np.uint8)
    mask[crow-30:crow+30, ccol-30:ccol+30] = 1

    # 掩膜图像和频谱图像乘积
    f = fshift * mask

    # 傅里叶逆变换
    ishift = np.fft.ifftshift(f)
    iimg = cv.idft(ishift)
    res = cv.magnitude(iimg[:,:,0], iimg[:,:,1])
    
    return res
# 高通滤波
def High_Pass_Filter(srcImg_path):
    #img = cv.imread(srcImg_path, 0)
    img = np.array(Image.open(srcImg_path))
    img = cv.cvtColor(img,cv.COLOR_BGR2GRAY)

    # 傅里叶变换
    dft = cv.dft(np.float32(img), flags = cv.DFT_COMPLEX_OUTPUT)
    fshift = np.fft.fftshift(dft)

    # 设置高通滤波器
    rows, cols = img.shape
    crow, ccol = int(rows/2), int(cols/2) # 中心位置
    mask = np.ones((rows, cols, 2), np.uint8)
    mask[crow-30:crow+30, ccol-30:ccol+30] = 0

    # 掩膜图像和频谱图像乘积
    f = fshift * mask

    # 傅里叶逆变换
    ishift = np.fft.ifftshift(f)
    iimg = cv.idft(ishift)
    res = cv.magnitude(iimg[:,:,0], iimg[:,:,1])
    return res
# 均值滤波
img = cv.imread('./CopyRNeRF-code/data/r_0.png')
# img_blur = cv.blur(img, (3,3)) # (3,3)代表卷积核尺寸，随着尺寸变大，图像会越来越模糊
# img_blur=cv.bilateralFilter(img, 50, 100, 100)
img_blur= cv.GaussianBlur(img, (3,3), 0, 0)

# img_blur = cv.cvtColor(img_blur, cv.COLOR_BGR2RGB) # BGR转化为RGB格式
cv.imwrite('./data/t.png',img_blur)
# img_High_Pass_Filter =High_Pass_Filter('./data/r_0.png')
# imageio.imwrite('./data/t.png',img_High_Pass_Filter*255)