I had posted a question previously about the Richardson-Lucy algorithm. I have a follow-up question I would appreciate help with.
Below is the Python code I am using. My input image is already blurry so I removed program lines that I originally had to intentionally blur the image. I am getting the error "RuntimeWarning: invalid value encountered in true_divide relative_blur = image / convolve(im_deconv, psf, mode='same')" I would appreciate help with debugging this. I kept the lines in the program that I commented out based on the suggestion below.
import numpy as np
import matplotlib.pyplot as plt
from PIL import Image, ImageFilter
from scipy.signal import convolve2d as conv2
from skimage import color, data, restoration
Image.open('TOFA-003_UV_Cured_Lincoln_Corrected.bmp').convert('L').save('TOFA-003_UV_Cured_Lincoln_Corrected_gray.bmp')
astro = Image.open('TOFA-003_UV_Cured_Lincoln_Corrected_gray.bmp')
psf = np.ones((5, 5)) / 25
#psf = np.ones((8, 8)) / 25
astro = conv2(astro, psf, 'same')
astro = astro/255
# Add Noise to Image
#astro_noisy = astro.copy()
#astro_noisy += (np.random.poisson(lam=25, size=astro.shape) - 10) / 255
#astro_noisy = astro_noisy/255
# Restore Image using Richardson-Lucy algorithm
deconvolved_RL = restoration.richardson_lucy(astro, psf, iterations=2)
fig, ax = plt.subplots(nrows=1, ncols=3, figsize=(8, 5))
plt.gray()
for a in (ax[0], ax[1], ax[2]):
a.axis('off')
ax[0].imshow(astro)
ax[0].set_title('Original Data')
#ax[1].imshow(astro_noisy)
#ax[1].set_title('Noisy data')
ax[2].imshow(deconvolved_RL, vmin=astro.min(), vmax=astro.max())
ax[2].set_title('Restoration using\nRichardson-Lucy')
fig.subplots_adjust(wspace=0.02, hspace=0.2,
top=0.9, bottom=0.05, left=0, right=1)
plt.show()
Related
I am new to Python and am trying to modify an existing Richardson-Lucy program for an image that I have.
Specifically, I have a bitmap image 'flower2.bmp' that I am using to test the following program:
import numpy as np
import matplotlib.pyplot as plt
from PIL import Image, ImageFilter
from scipy.signal import convolve2d as conv2
from skimage import color, data, restoration
Image.open('flower2.bmp').convert('L').save('flower2_gray.bmp')
astro = Image.open('flower2_gray.bmp')
psf = np.ones((5, 5)) / 25
astro = conv2(astro, psf, 'same')
# Add Noise to Image
astro_noisy = astro.copy()
astro_noisy += (np.random.poisson(lam=25, size=astro.shape) - 10) / 255.
# Restore Image using Richardson-Lucy algorithm
deconvolved_RL = restoration.richardson_lucy(astro_noisy, psf, iterations=100)
fig, ax = plt.subplots(nrows=1, ncols=3, figsize=(8, 5))
plt.gray()
for a in (ax[0], ax[1], ax[2]):
a.axis('off')
ax[0].imshow(astro)
ax[0].set_title('Original Data')
ax[1].imshow(astro_noisy)
ax[1].set_title('Noisy data')
ax[2].imshow(deconvolved_RL, vmin=astro_noisy.min(), vmax=astro_noisy.max())
ax[2].set_title('Restoration using\nRichardson-Lucy')
fig.subplots_adjust(wspace=0.02, hspace=0.2,
top=0.9, bottom=0.05, left=0, right=1)
plt.show()
Below is the image output. I would appreciate help understanding why the Restoration image is black.
Thank you.
The restoration.richardson_lucy documentation notes that the function has an optional "clip" argument, which is true by default:
clip : boolean, optional
True by default. If true, pixel value of the result above 1 or under -1 are thresholded for skimage pipeline compatibility.
However, the astro image read from Image.open has nominal range [0, 255]. I'm guessing that all pixels in astro are >= 1, so that the result after clipping to [-1, 1] is simply a constant image of ones. This could explain the plot.
The fix is to divide astro_noisy by 255 before passing it to restoration.richardson_lucy.
my problem is that I always get the following error when I operate the following code.Strange thing is that, when i set the epochs to 0 the error dosnt show up and I can upload with no problems. Thanks for the Help!
I have already tried anabling third party cockies, which did not help. The strange thing is, that the upload works, if I set the training epochs to 0.
Sometimes the error is google.colab._files is undefined.
I have already tried to use Chrome and Firefox.
import tensorflow as tf
import numpy as np
mnist = tf.keras.datasets.fashion_mnist
(training_images, training_labels), (test_images, test_labels) = mnist.load_data()
training_images = training_images.reshape(60000, 28, 28, 1)
training_images = training_images / 255.0
test_images = test_images.reshape(10000, 28, 28, 1)
test_images = test_images / 255.0
model = tf.keras.models.Sequential([
tf.keras.layers.Conv2D(32,(3,3),activation='relu', input_shape=(28,28,1)),
tf.keras.layers.MaxPooling2D(2,2),
tf.keras.layers.Conv2D(64,(3,3),activation='relu'),
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(128, activation='relu'),
tf.keras.layers.Dense(10, activation='softmax')
])
model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.fit(training_images,training_labels, epochs=1)
classes = model.predict(test_images)
predicted_classes = np.argmax(classes, axis=1)
print(classes[0])
print(test_labels[0])
mnist = tf.keras.datasets.fashion_mnist
(training_images, training_labels), (test_images, test_labels) = mnist.load_data()
import matplotlib.pyplot as plt
plt.imshow(test_images[0], cmap='Greys_r')
import numpy as np
from google.colab import files
from keras.preprocessing import image
import cv2
import matplotlib.pyplot as plt
uploaded = files.upload()
for fn in uploaded.keys():
path = '/content/' + fn
img = cv2.imread(path)
img = cv2.resize(img,(28,28))
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
x = image.img_to_array(img, dtype=np.float32)
print("top left pixel value:", x[0,0])
if x[0,0] > 250:
# white background
print("needs to be inverted!")
x -= 255
x *= -1
x = x / 255.0
x = x.reshape(1, 28, 28, 1)
plt.imshow(img, cmap='Greys_r')
plt.show()
classes = model.predict(x)
plt.bar(range(10), classes[0])
plt.show()
print("prediction: class", np.argmax(classes[0]))
TypeError: Cannot read property '_uploadFiles' of undefined
So I found out that it works if you use 2 cells one for the neural network and one for the upload feature.
PyTorch newbie here. I wrote a script (code below) that performs the following operations: load an image, perform a 2D convolution operation and then display the output and the input.
At present I have the image below, which seems off. How can I plot the feature map correctly?
import numpy as np
import torch
import torchvision
import torchvision.transforms as transforms
import torch.nn as nn
import matplotlib.pyplot as plt
import imageio
import sys
A = imageio.imread('LiT.png')
# Define how the convolution operation works
conv2 = nn.Conv2d(in_channels=3, out_channels=3, kernel_size=3, stride=1, padding=1)
image_d = torch.FloatTensor(np.asarray(A.reshape(1, 3, A.shape[0] , A.shape[1])))
fc = conv2(image_d)
fc1 = fc.permute(0, 2, 3, 1).reshape([516, 780, 3])
plt.figure(figsize=(16,8))
plt.subplot(1,2,1)
plt.imshow(A)
plt.subplot(1,2,2)
plt.imshow(fc1.data.numpy())
plt.show()
The issue with your code is this line
image_d = torch.FloatTensor(np.asarray(A.reshape(1, 3, A.shape[0] , A.shape[1])))
You can't just reshape the image you need to transpose the channels. As a remark for the future, if you get a stripy result like you did it's most likely some permutation/transposition or reshaping operation that's not correct.
Other than that I also scaled the input image to [0, 1] to show it properly. Below is the working code:
import numpy as np
import torch
import torchvision
import torchvision.transforms as transforms
import torch.nn as nn
import matplotlib.pyplot as plt
import imageio
import sys
A = imageio.imread('LiT.png')
# Define how the convolution operation works
conv2 = nn.Conv2d(in_channels=3, out_channels=3, kernel_size=3, stride=1, padding=1)
# from [H, W, C] to [C, H, W]
transposed_image = A.transpose((2, 0, 1))
# add batch dim
transposed_image = np.expand_dims(transposed_image, 0)
image_d = torch.FloatTensor(transposed_image)
fc = conv2(image_d)
fc1 = fc.permute(0, 2, 3, 1)[0]
result = fc1.data.numpy()
max_ = np.max(result)
min_ = np.min(result)
result -= min_
result /= max_
plt.figure(figsize=(16,8))
plt.subplot(1,2,1)
plt.imshow(A)
plt.subplot(1,2,2)
plt.imshow(result)
plt.show()
To my understanding, the problem lies in how you are permuting channels position in the image by using reshape. Instead, 'np.transpose or tensor.permute should be used. Using torch for permutation:
image_d = torch.FloatTensor(np.asarray(A)).unsqueeze(0).permute(0,3,1,2)
Or, if we want to handle the permutation part in numpy:
image_d = np.transpose(np.asarray(A), (2,0,1))
image_d = torch.FloatTensor(image_d).unsqueeze(0)
I want to make program checking my foot-size
I don't know this error about opencv
url: http://cocoding94.blogspot.com/2017/05/blog-post_7.html .
import numpy as np
import matplotlib.pyplot as plt
import cv2
img = cv2.imread("foot.jpeg")
blur = cv2.blur(img,(5,10))
rows,cols,ch = img.shape
pts1 = np.float32([170,270],[480,220],[240, 710],[540,650])
pts2 = np.float32([0,0],[210,0],[0,297],[210,297])
M = cv2.getPerspectiveTransform(pts1,pts2)
dst = cv2.warpPerspective(img,M,(210,297))
plt.subplot(121),plt.imshow(img),plt.title('Input')
plt.plot(*zip(*point), marker='.', color='r', ls='')
plt.subplot(122),plt.imshow(dst),plt.title('Output')
plt.show()
In raspberry pi error printing:
Traceback (most recent call last): File "foot.py",line 7,in
blur = cv2. blur (img,(5,10)) cv2.error:OpenCV(3.4.3)
/home/pi/opencv/opencv-3.4.3/modules/core/src/matrix.cpp:756: error:
(-215:Assertion failed) dims <=2 && step[0] > 0 in function
'locateROI'
but please next time add more information to make this community greater than another, I fixed some errors in your code, i don´t your original image, but the code now works well, try to change your point to get a good perspective:
import numpy as np
import matplotlib.pyplot as plt
import cv2
img = cv2.imread("machupichu.jpg")
#blur = cv2.blur(img,(5,5))
#rows,cols,ch = img.shape
point=[[170,270],[480,220],[240, 710],[540,650]]
pts1 = np.float32([[170,270],[480,220],[240, 710],[540,650]])
pts2 = np.float32([[0,0],[210,0],[0,297],[210,297]])
M = cv2.getPerspectiveTransform(pts1,pts2)
dst = cv2.warpPerspective(img,M,(210,297))
plt.subplot(121)
plt.imshow(img)
plt.title('Input')
plt.plot(*zip(*point), marker='.', color='r', ls='')
plt.subplot(122)
plt.imshow(dst)
plt.title('Output')
plt.show()
Here is the image which I used:
Machu Pichu
This is the result:
Best Regards.
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt
from scipy.misc import imread
img = imread('dog2.jpg')
#img is a shape of (360, 480, 3)
w = img.shape[0]
h = img.shape[1]
c = img.shape[2]
k = 3 # for my convenience
plt.subplot(1,2,1)
plt.imshow(img)
img = tf.cast(img, tf.float32)
img4d = tf.reshape(img,[1,w,h,c])
diag = np.array([[1,1,1],[0,0,0],[1,1,1]]*k, np.float32)
# diag = np.diag(diag)
diag4d = tf.reshape(diag,[k,k,c,1])
convolved = tf.nn.conv2d(img4d, diag4d, strides=[1,1,1,1], padding='SAME')
with tf.Session() as sess:
result = sess.run(convolved)
print result.shape
plt.subplot(1,2,2)
plt.imshow(np.squeeze(result))
plt.show()
I am just trying to use convolution and apply some blur effect initially. Yeah I know that my kernel values aren't right. But my question is, I am giving an input image that has 3 channels. How could I get an output image of 3 channels. Well. I tried. But all I get is some one channeled values alone.
You are passing a kernel of shape [3, 3, 3, 1] to tf.nn.conv2d(). If you want to get a 3-channel image output from your convolution, the fourth dimension of your kernel (called out_channels in the official documentation) should be 3 instead of 1; [3, 3, 3, 3] for example.
You could also take a look at the conv2d documentation, this question and this question to better understand Tensorflow's conv2d method.