Cannot open file (/var/www/vhosts/aprendtech.com/httpdocs/wordpress/wp-content/backup/.htaccess)Cannot write to file (/var/www/vhosts/aprendtech.com/httpdocs/wordpress/wp-content/backup/.htaccess) AprendBlog » Implementation

Jul 25 2011

Matlab interface to the projection simulator

Tag: Implementation,Physics,software — admin @ 3:45 pm

edit: Oct. 4, 2011 The C++ code referred to in this post has been superseded by a Matlab only version that is faster and has more geometric shapes. The Matlab function uses the approach described here so it gives similar results.
My last post discussed C++ code for a CT projection simulator. In this post, I will describe a Matlab interface to the simulator executable.

more –>;

Jul 15 2011

A projection simulator for image reconstruction research–C++ implementation

Tag: Implementation,software — admin @ 7:08 pm

edit: Oct. 4, 2011 The C++ code referred to in this post has been superseded by a Matlab only version that is faster and has more geometric shapes. The Matlab function uses the approach described here so it gives similar results.
My last post discussed the approach for my CT projection simulator. In this post, I will describe the C++ implementation.

 

more –>;

Jul 06 2011

A projection simulator for image reconstruction research–rationale

Tag: Implementation,Physics,software — admin @ 9:53 am

edit: Oct. 4, 2011 The C++ code referred to at the end of this post has been superseded by a Matlab only version that is faster and has more geometric shapes. The Matlab function uses the approach described here so the rationale is still valid.
Simulators are essential tools in CT research. They provide data with known properties to test reconstruction algorithms and simulations of complete systems. Matlab provides simulators with the radon and fanbeam functions of the Image Processing Toolbox. However, these functions specify the object as a raster image, which leads to large artifacts in the projections and in the reconstructed images.

more –>;

Jun 18 2011

An improved parallel projection image reconstructor for Matlab

Tag: Implementation,software — admin @ 1:18 pm

This post describes a replacement for iradon for Matlab and Octave. The new function, CTrecon.m,

1. Fixes a bug in iradon that results in offsets from the true values in the reconstructed image.

2. Uses a mex function from iradon_speedy for backprojection. According to a comment by Jeff Orchard, the author of iradon_speedy, on the Mathworks File Exchange entry for iradon_speedy, Matlab R2011a now uses a mex function. I do not have that version so I cannot confirm this.

3. Handles complex projections to return a complex reconstructed image – see zBackproject.c

4. Regularizes the interface to (parameter-name’,parameter-value) syntax. This implies that CTrecon cannot just be dropped in for iradon. The advantage is more readable code.

more –>;

« Previous Page