|
|
|
|
|
import sys
|
|
|
|
|
|
import random
|
|
|
|
|
|
import cv2
|
|
|
|
|
|
import numpy as np
|
|
|
|
|
|
import matplotlib.pyplot as plt
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def addNoisy(img_path, type, para1, para2):
|
|
|
|
|
|
img = cv2.imread(img_path) # 读取图片
|
|
|
|
|
|
out = np.zeros(img.shape, np.uint8)
|
|
|
|
|
|
if type == 1:
|
|
|
|
|
|
# 添加高斯噪声-------------------------------------------------------
|
|
|
|
|
|
# 将图片的像素值归一化，存入矩阵中
|
|
|
|
|
|
img = np.array(img / 255, dtype=float)
|
|
|
|
|
|
# 生成正态分布的噪声，其中0表示均值，0.1表示方差
|
|
|
|
|
|
noise = np.random.normal(para1, para2, img.shape) # 参数：正态分布的均值和方差，可自定义
|
|
|
|
|
|
# 将噪声叠加到图片上
|
|
|
|
|
|
out = img + noise
|
|
|
|
|
|
# 将图像的归一化像素值控制在0和1之间，防止噪声越界
|
|
|
|
|
|
out = np.clip(out, 0.0, 1.0)
|
|
|
|
|
|
# 将图像的像素值恢复到0到255之间
|
|
|
|
|
|
out = np.uint8(out * 255)
|
|
|
|
|
|
cv2.imwrite('./output/Gaussian_noisy.jpg', out)
|
|
|
|
|
|
elif type == 2:
|
|
|
|
|
|
# 添加椒盐噪声--------------------------------------------------------
|
|
|
|
|
|
# 遍历图像，获取叠加噪声后的图像
|
|
|
|
|
|
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
|
|
|
|
|
|
for i in range(img.shape[0]):
|
|
|
|
|
|
for j in range(img.shape[1]):
|
|
|
|
|
|
rdn = random.random()
|
|
|
|
|
|
if rdn < para1:
|
|
|
|
|
|
# 添加椒噪声
|
|
|
|
|
|
out[i][j] = 0
|
|
|
|
|
|
elif rdn > para2:
|
|
|
|
|
|
# 添加盐噪声
|
|
|
|
|
|
out[i][j] = 255
|
|
|
|
|
|
else:
|
|
|
|
|
|
# 不添加噪声
|
|
|
|
|
|
out[i][j] = img[i][j]
|
|
|
|
|
|
cv2.imwrite('./output/saltnoisy.jpg', out)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
'''
|
|
|
|
|
|
输入：img_path:图片地址
|
|
|
|
|
|
type：噪声类型(1为高斯噪声，2为椒盐噪声）
|
|
|
|
|
|
para1,para2：如果添加的是高斯噪声，para1为正态分布均值，para2为正态分布方差
|
|
|
|
|
|
如果添加的是椒盐噪声，para1为添加椒噪声阈值，para2为添加盐噪声阈值
|
|
|
|
|
|
|
|
|
|
|
|
输出：添加了噪声的图片
|
|
|
|
|
|
'''
|
|
|
|
|
|
if __name__ == '__main__':
|
|
|
|
|
|
addNoisy(sys.argv[1], eval(sys.argv[2]), eval(sys.argv[3]), eval(sys.argv[4]))
|
