Image Reconstruction Using Filtered Back Projection Algorithm

Image Reconstruction Using Filtered Back Projection Algorithm

Image reconstruction is one of the important links in digital image processing, and filtered back projection algorithm is one of the commonly used methods. This article mainly introduces how to use MATLAB to implement the filtered back projection algorithm to reconstruct the input image.

First, we need to prepare a CT (Computed Tomography) scan image to be reconstructed. Suppose we have got a CT scan image named "ct_scan.jpg", we can use the imread function to read it into MATLAB:

CT_scan = imread('ct_scan.jpg');

After reading the image, we also need to preprocess the image, including converting the image to a grayscale image, normalizing and other steps. Here we use im2gray and mat2gray functions to implement these two steps respectively:

gray_scan = im2gray(CT_scan);
norm_scan = mat2gray(double(gray_scan));

Next, we need to define some parameters related to the algorithm, including the scanning angle range, the number of sampling points, the filter type and the size of the reconstructed image, etc. Here we assume that the scanning angle range is 180 degrees, the number of sampling points is 360, the filter type is Ram-Lak filter, and the reconstructed image size is 256 x 256 pixels. Then we can define these parameters as follows:

angle_range = 180;
num_samples = 360;
filter_type = 'ram-lak';
recon_size = [256, 256];

Next comes the core part of the algorithm, the filtered back-projection process. Here we use the iradon function for backprojection, the ramp function to create a Ram-Lak filter, and the radon function for raw projection. whole