Image Processing Codes
1:
Read Lena image from the disk and display the image in figure window. Write the image on the
different path on the disk.
2:
Read two images I1 and I2 of your choice and do the following without using any inbuilt
function:
3:
Read Lena image and check whether the image is 4-Connected, 8- Connected or m- Connected.
4:
A black and white input image I(x,y) which gray level vary from (0 to L) and after point transformation output intensity vary from (0 to L). where u,v is the input / output image intensity gray level. Show output for these case
V = f( u )
Case-1 Output image are negative.
Case-2 Output image remove highly dark and white region.
Case-3 Output image remove background, background region intensity between [100, 150].
Case-4 Non linear stretching (logarithmic).
5:
Given cameraman.tif, use gray level slicing to,
I. Separate camera man from background.
II. Brighten the camera man while preserving the background.
III. Plot the graph between input and output intensities for both the cases.
6:
Plot the histogram for pout.tif and do the histogram equalization without using histeq command.
I. Find the mean and variance of the image.
II. Do the local histogram equalization using a 4*4 neighborhood with and without overlapping.
7:
Given an image “embedded_square_noisy.tif” perform local histogram equalization. Also perform global histogram equalization and show the differences visually.
Note: You can use built-in functions only for this question.
8:
Generate a Gaussian filter mask of sizes m=3,5,7,9 and varying σ . Perform smoothening on the given image. Show the images visually using subplot with titles.
9:
You are given a noisy image and you only know two filtering techniques average filtering and median filtering. Apply those operations on the noisy image and show the differences.
10:
You are asked to sharpen given blurry image using four different Laplacian masks.
11:
Given the image perform unsharp masking and high boost filtering. Show each steps visually.
12:
Take an image of your choice. Use the laplacian filter to show the implementation of unsharp masking and High boost filtering in the image both in spatial domain and frequency domain. Are the two outputs same? If yes/no, explain the reason of it.
13:
Implement high frequency emphasis filtering using any high pass filtering. Show the effect on the image.
14:
Implement homomorphic filtering to show its effects on an image of your choice.
15:
Given an imagef(x, y), your task is to degrade it using the degradation function H(u, v) given as –
H(u, v) = (T/π(ua + vb))*sinπ(ua + vb)))
Assume T = 1, a = b = 0.1
Hint: Compute the Fourier transform F(u, v) of f(x, y), then G(u, v) = H(u, v)F(u, v). Finally g(x, y) = I^(−1){G(u, v)}
Now that you know the degradation function, your next task is to recover the degraded image g(x, y) to approximate the given original image f(x, y).
16:
Observing the given image f(x, y), one can see horizontal scan lines. Your task here is to remove those horizontal lines to get improved image f′(x, y). This has to be done using notch reject filering. Display the result and the extracted noise pattern.
Hint: Involves Fourier transform operations.
