本文介绍了向图像添加高斯噪声的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!
问题描述
我正在尝试使用以下代码向某些图像添加高斯噪声
I'm trying to add gaussian noise to some images using the following code
import numpy as np
import cv2
import glob
mean = 0
var = 10
sigma = var ** 0.5
gaussian = np.random.normal(mean, sigma, (224, 224))
for image in glob.glob('/home/aub/myflower/flower_photos/dandelion/*.jpg'):
img = cv2.imread(image)
noisy_image = np.zeros(img.shape, np.float32)
if len(img.shape) == 2:
noisy_image = img + gaussian
else:
noisy_image[:, :, 0] = img[:, :, 0] + gaussian
noisy_image[:, :, 1] = img[:, :, 1] + gaussian
noisy_image[:, :, 2] = img[:, :, 2] + gaussian
cv2.normalize(noisy_image, noisy_image, 0, 255, cv2.NORM_MINMAX, dtype=-1)
noisy_image = noisy_image.astype(np.uint8)
cv2.imshow("img", img)
cv2.imshow("gaussian", gaussian)
cv2.imshow("noisy", noisy_image)
cv2.waitKey(0)
但它不起作用,它给了我以下错误
but it doesn't work and it gives me the following error
noisy_image[:, :, 0] = img[:, :, 0] + 高斯值错误:操作数无法与形状一起广播 (315,500) (224,224)
请查看并提供反馈.
推荐答案
看起来你的图像形状是(315,500)
,而gaussian
的形状是(224,224)
.尝试将您的高斯初始化更改为
It looks like your image shape is (315,500)
, while the shape of gaussian
is (224,224)
. Try changing your gaussian initialization to
gaussian = np.random.normal(mean, sigma, (img.shape[0],img.shape[1]))
顺便说一句:您可以替换这些行
By the way:You can replace these lines
noisy_image[:, :, 0] = img[:, :, 0] + gaussian
noisy_image[:, :, 1] = img[:, :, 1] + gaussian
noisy_image[:, :, 2] = img[:, :, 2] + gaussian
与
noisy_image = img + gaussian
这将产生相同的效果:将 gaussian
添加到每个通道.
which will have the same effect: adding gaussian
to each channel.
这篇关于向图像添加高斯噪声的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持!