数字图像噪声_Python
一：数字图像噪声来源
图像的获取（数字化过程）和传输过程，当使用相机获取图像时，光照程度和传感器温度是生成图像大量噪点的主要因素。
二：灰度图加入高斯噪声（正态噪声）
def GaussianNoise(src,mean,sigma):
NoiseImg=src
rows,cols=NoiseImg.shape
for i in range(rows):
for j in range(cols):
NoiseImg[i,j]=NoiseImg[i,j]+random.gauss(mean,sigma)
if NoiseImg[i,j]<0:
NoiseImg[i,j]=0
elif NoiseImg[i,j]>255:
NoiseImg[i,j]=255
return NoiseImg
img=cv2.imread('IMG_4470.JPG',0)
img=GaussianNoise(img,2,4)
cv2.imwrite('IMG_4470_GaussianNoise.jpg',img)
cv2.imshow('IMG_4470_GaussianNoise',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
cv2.waitKey(1)
输出结果为：
原图为：
三：彩色图加入高斯噪声
def GaussianNoise_color(src,mean,sigma):
NoiseImg=src
rows,cols,_=NoiseImg.shape
for i in range(rows):
for j in range(cols):
NoiseImg[i,j]=NoiseImg[i,j]+random.gauss(mean,sigma)
return NoiseImg
img=cv2.imread('IMG_4470.JPG')
img=GaussianNoise_color_percentage(img,2,4)
cv2.imwrite('IMG_4470_GaussianNoise_color_1.jpg',img)
cv2.imshow('IMG_4470_GaussianNoise_color_1',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
cv2.waitKey(1)
输出为：
四：部分像素点加入高斯噪声
def GaussianNoise_color_percentage(src,mean,sigma,percentage):
NoiseImg=src
rows,cols,_=NoiseImg.shape
num=int(rows*cols*percentage)
for i in range(num):
randX=np.random.randint(0,rows-1)
randY=np.random.randint(0,cols-1)
NoiseImg[randX,randY]=NoiseImg[randX,randY]+random.gauss(mean,sigma)
return NoiseImg
img=cv2.imread('IMG_4470.JPG')
img=GaussianNoise_color_percentage(img,2,4,0.5)
cv2.imwrite('IMG_4470_GaussianNoise_color_percentage_1.jpg',img)
cv2.imshow('IMG_4470_GaussianNoise_color_percentage_1',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
cv2.waitKey(1)
输出结果为：
五：图像加入椒盐噪声_胡椒点
def Pepper(src,percentage):
NoiseImg=src
rows,cols,_=NoiseImg.shape
NoiseNum=int(percentage*rows*cols)
for i in range(NoiseNum):
randX=np.random.randint(0,rows-1)
randY=np.random.randint(0,cols-1)
if random.randint(0,1)<=0.5:
NoiseImg[randX,randY]=0
else:
NoiseImg[randX,randY]=NoiseImg[randX,randY]
return NoiseImg
img=cv2.imread('IMG_4470.JPG')
img=PepperandSalt_2(img,0.01)
cv2.imwrite('IMG_4470_PepperandSalt_7_pepper.jpg',img)
cv2.imshow('IMG_4470_PepperandSalt_7',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
cv2.waitKey(1)
输出结果为：
六：图像加入椒盐噪声_盐点
def salt(src,percentage):
NoiseImg=src
rows,cols,_=NoiseImg.shape
NoiseNum=int(percentage*rows*cols)
for i in range(NoiseNum):
randX=np.random.randint(0,rows-1)
randY=np.random.randint(0,cols-1)
if random.randint(0,1)<=0.5:
NoiseImg[randX,randY]=255
else:
NoiseImg[randX,randY]=NoiseImg[randX,randY]
return NoiseImg
img=cv2.imread('IMG_4470.JPG')
img=salt(img,0.01)
cv2.imwrite('IMG_4470_PepperandSalt_8_salt.jpg',img)
cv2.imshow('IMG_4470_PepperandSalt_8',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
cv2.waitKey(1)
输出结果为:
七：灰度图加入椒盐噪声
def PepperandSalt(src,percentage):
NoiseImg=src
rows,cols=NoiseImg.shape
NoiseNum=int(percentage*rows*cols)
for i in range(NoiseNum):
randX=np.random.randint(0,rows-1)
randY=np.random.randint(0,cols-1)
if random.randint(0,1)<=0.5:
NoiseImg[randX,randY]=0
else:
NoiseImg[randX,randY]=255
return NoiseImg
img=cv2.imread('IMG_4470.JPG',0)
img=PepperandSalt(img,0.05)
cv2.imwrite('IMG_4470_PepperandSalt_2.jpg',img)
cv2.imshow('IMG_4470_PepperandSalt_2',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
cv2.waitKey(1)
输出结果为：
八：彩色图加入椒盐噪声
def PepperandSalt_color(src,percentage):
NoiseImg=src
rows,cols,_=NoiseImg.shape
NoiseNum=int(percentage*rows*cols)
for i in range(NoiseNum):
randX=np.random.randint(0,rows-1)
randY=np.random.randint(0,cols-1)
if random.randint(0,1)<=0.5:
NoiseImg[randX,randY]=0
else:
NoiseImg[randX,randY]=255
return NoiseImg
img=cv2.imread('IMG_4470.JPG')
img=PepperandSalt_color(img,0.05)
cv2.imwrite('IMG_4470_PepperandSalt_color.jpg',img)
cv2.imshow('IMG_4470_PepperandSalt_color',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
cv2.waitKey(1)
输出结果为：
九：指数分布噪声
def Exponent(src,constant,percentage):
NoiseImg=src
rows,cols=NoiseImg.shape
num=int(rows*cols*percentage)
for i in range(num):
randX=np.random.randint(0,rows-1)
randY=np.random.randint(0,cols-1)
if NoiseImg[randX,randY]>128:
NoiseImg[randX,randY]=constant*np.exp(-constant*NoiseImg[randX,randY])
else:
NoiseImg[randX,randY]=NoiseImg[randX,randY]
return NoiseImg
img=cv2.imread('IMG_4470.JPG',0)
img=Exponent(img,0.5,0.01)
cv2.imwrite('IMG_4470_Exponent.jpg',img)
cv2.imshow('IMG_4470_Exponent',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
cv2.waitKey(1)
输出结果为：
十：均匀分布噪声
def distribution(src,low,high,percentage):
NoiseImg=src
rows,cols=NoiseImg.shape
num=int(rows*cols*percentage)
for i in range(num):
randX=np.random.randint(0,rows-1)
randY=np.random.randint(0,cols-1)
if low输出结果为：
十一：伽马（爱尔兰）噪声
伽马曲线如下：
a=10
b=20
z=np.arange(1,20,0.1)
y=(a**b)*(z**(b-1))/np.math.factorial(b-1)*np.exp(-a*z)
plt.plot(z,y)
def gamma_noise(src,constant_a,constant_b,percentage):
NoiseImg=src
rows,cols=NoiseImg.shape
num=int(rows*cols*percentage)
for i in range(num):
randX=np.random.randint(0,rows-1)
randY=np.random.randint(0,cols-1)
if NoiseImg[randX,randY]>120 and b!=0:
z=NoiseImg[randX,randY]
NoiseImg[randX,randY]=(a**b)*(z**(b-1))/np.math.factorial(b-1)*np.exp(-a*z)
else:
NoiseImg[randX,randY]=NoiseImg[randX,randY]
return NoiseImg
img=cv2.imread('IMG_4470.JPG',0)
img=gamma_noise(img,5,10,0.01)
cv2.imwrite('IMG_4470_gamma_noise.jpg',img)
cv2.imshow('IMG_4470_gamma_noise',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
cv2.waitKey(1)
输出结果为：
