Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: THEend8_
RMIT Medical Physics
PHYS2134
Advanced Medical Imaging
ASSIGNMENT: Image Processing & Reconstruction
Question 1
Open the angiogram image in ImageJ. Select as large a region as possible with near-uniform grey
values, using the polygon selection tool. Find the signal-to-noise ratio.
Smooth the image ( Process/Smooth ). How does this affect the shape of the histogram? How do the
mean and standard deviation change? Try several iterations of smoothing to see the trend clearly.
Now sharpen the original image ( Process/ Sharpen ). How does this affect the shape of the
histogram? How do the mean and standard deviation (and dynamic range) change? Again, try
several iterations to see the trend clearly.
Explain the changes you observe after smoothing and sharpening, supported by your data/evidence.
Add Gaussian noise ( Process/Noise/Add Specified Noise ) with a standard deviation of 10 to the
original angiogram image. Compare the updated histogram with the histogram of the original image,
in terms of its overall shape, its standard deviation and its dynamic range. Estimate the signal-to-
noise ratio of the noisy image. Repeat after adding noise of standard deviation 20 to the original
image. How does the signal-to-noise ratio change with added noise?
Note: Please include the screen shots of Histograms while explaining the changes in histograms.
Hint: Use live histogram
Question 2
Open noisyS&Pskull, which is a radiograph of a skull contaminated by salt-and-pepper noise.
Use the live histogram plugin to display its grey-level histogram: note the large spikes at 0 and 255…
What do these spikes represent? Apply a 5 × 5 average mask using Process/Filters … Mean with a
radius of 2. Note how the histogram changes. Compare the result with that of using a 5 × 5 median
mask ( Process/ Filters … Median with a radius of 2 on the original noisy image.
Which is more successful at removing this kind of noise and why? How do the histograms of the
resulting images differ?