滤波算法

111222333 · 2022-06-11 16:30:03

import matplotlib.pyplot as plt
import numpy as np
from numpy import fft
import math
import cv2

# 仿真运动模糊
def motion_process(image_size, motion_angle):
PSF = np.zeros(image_size)
print(image_size)
center_position = (image_size[0] - 1) / 2
print(center_position)

slope_tan = math.tan(motion_angle * math.pi / 180)
slope_cot = 1 / slope_tan
if slope_tan <= 1:
for i in range(15):
offset = round(i * slope_tan) # ((center_position-i)*slope_tan)
PSF[int(center_position + offset), int(center_position - offset)] = 1
return PSF / PSF.sum() # 对点扩散函数进行归一化亮度
else:
for i in range(15):
offset = round(i * slope_cot)
PSF[int(center_position - offset), int(center_position + offset)] = 1
return PSF / PSF.sum()

# 对图片进行运动模糊
def make_blurred(input, PSF, eps):
input_fft = fft.fft2(input) # 进行二维数组的傅里叶变换
PSF_fft = fft.fft2(PSF) + eps
blurred = fft.ifft2(input_fft * PSF_fft)
blurred = np.abs(fft.fftshift(blurred))
return blurred

def inverse(input, PSF, eps): # 逆滤波
input_fft = fft.fft2(input)
PSF_fft = fft.fft2(PSF) + eps # 噪声功率，这是已知的，考虑epsilon
result = fft.ifft2(input_fft / PSF_fft) # 计算F(u,v)的傅里叶反变换
result = np.abs(fft.fftshift(result))
return result

def wiener(input, PSF, eps, K=0.01): # 维纳滤波，K=0.01
input_fft = fft.fft2(input)
PSF_fft = fft.fft2(PSF) + eps
PSF_fft_1 = np.conj(PSF_fft) / (np.abs(PSF_fft) ** 2 + K)
result = fft.ifft2(input_fft * PSF_fft_1)
result = np.abs(fft.fftshift(result))
return result

def normal(array):
array = np.where(array < 0, 0, array)
array = np.where(array > 255, 255, array)
array = array.astype(np.int16)
return array

def main(gray):
channel = []
img_h, img_w = gray.shape[:2]
PSF = motion_process((img_h, img_w), 60) # 进行运动模糊处理
blurred = np.abs(make_blurred(gray, PSF, 1e-3))

result_blurred = inverse(blurred, PSF, 1e-3) # 逆滤波
result_wiener = wiener(blurred, PSF, 1e-3) # 维纳滤波

blurred_noisy = blurred + 0.1 * blurred.std() *
np.random.standard_normal(blurred.shape) # 添加噪声,standard_normal产生随机的函数
inverse_mo2no = inverse(blurred_noisy, PSF, 0.1 + 1e-3) # 对添加噪声的图像进行逆滤波
wiener_mo2no = wiener(blurred_noisy, PSF, 0.1 + 1e-3) # 对添加噪声的图像进行维纳滤波
channel.append((normal(blurred),normal(result_blurred),normal(result_wiener),
normal(blurred_noisy),normal(inverse_mo2no),normal(wiener_mo2no)))
return channel

if __name__ == '__main__':
image = cv2.imread('./gggg/001.png')
b_gray, g_gray, r_gray = cv2.split(image.copy())

Result = []
for gray in [b_gray, g_gray, r_gray]:
channel = main(gray)
Result.append(channel)
blurred = cv2.merge([Result[0][0][0], Result[1][0][0], Result[2][0][0]])
result_blurred = cv2.merge([Result[0][0][1], Result[1][0][1], Result[2][0][1]])
result_wiener = cv2.merge([Result[0][0][2], Result[1][0][2], Result[2][0][2]])
blurred_noisy = cv2.merge([Result[0][0][3], Result[1][0][3], Result[2][0][3]])
inverse_mo2no = cv2.merge([Result[0][0][4], Result[1][0][4], Result[2][0][4]])
wiener_mo2no = cv2.merge([Result[0][0][5], Result[1][0][5], Result[2][0][5]])

#========= 可视化 ==========
plt.figure(1)
plt.xlabel("Original Image")
plt.imshow(np.flip(image, axis=2)) # 显示原图像

plt.figure(2)
plt.figure(figsize=(8, 6.5))
imgNames = {"Motion blurred":blurred,
"inverse deblurred":result_blurred,
"wiener deblurred(k=0.01)":result_wiener,
"motion & noisy blurred":blurred_noisy,
"inverse_mo2no":inverse_mo2no,
'wiener_mo2no':wiener_mo2no}
for i,(key,imgName) in enumerate(imgNames.items()):
plt.subplot(231+i)
plt.xlabel(key)
plt.imshow(np.flip(imgName, axis=2))
plt.show()
OpenCV-Python 图像去模糊（维纳滤波，约束最小二乘方滤波）import matplotlib.pyplot as plt
import numpy as np
from numpy import fft
import math
import cv2