1887
banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
A method based on moving least squares for XRII image distortion correction
Rent:
Rent this article for
USD
10.1118/1.2791037
/content/aapm/journal/medphys/34/11/10.1118/1.2791037
http://aip.metastore.ingenta.com/content/aapm/journal/medphys/34/11/10.1118/1.2791037

Figures

Image of FIG. 1.
FIG. 1.

RMSD of MLS between true and corresponding positions in synthetic images in dependence on cutoff radius. Continuous lines refer to the Gauss weight function. Dotted lines refer to the exponential weight function. Dash-dotted lines refer to the rational formula weight function. Triangles indicate the optimal cutoff radii determined by the F-test (Ref. 12). Left plot shows the results obtained on the image of control grid. Right plot shows the results obtained on the image of control grid.

Image of FIG. 2.
FIG. 2.

RMSD of MLS between true and corresponding positions in synthetic images in dependence on basis function degree. Continuous line refers to the image of control grid. Dotted line refers to the image of control grid.

Image of FIG. 3.
FIG. 3.

RMSDs of five methods, L1, L2, GP, MLS, and MLSILS, as functions of number of control points. Thin continuous lines refer to MLS. Thick continuous lines refer to MLSILS. Thin dotted lines refer to GP. Thin dash-dotted lines refer to L1. Thick dash-dotted lines refer to L2. Left plot shows the RMSDs computed at control points. Right plot shows the RMSDs computed at intermediate points.

Image of FIG. 4.
FIG. 4.

Effect of sigmoidal distortion and pincushion distortion on RMSDs of the five correction methods, L1, L2, GP, MLS, and MLSILS. Thin continuous lines refer to MLS. Thick continuous lines refer to MLSILS. Thin dotted lines refer to GP. Thin dash-dotted lines refer to L1. Thick dash-dotted lines refer to L2. For the images of control grid, the RMSDs of L1 and L2 were very similar at intermediate points; the RMSDs of L2 was constant (zero) at control points over the whole range of distortions. For the sake of clarity, the RMSDs of L2 for the images of control grid are not plotted. For the images of control grid, the corresponding results were the same. The RMSDs of L2 and L1 are all plotted. At intermediate points, the RMSD lines of L1 are covered by that of L2.

Image of FIG. 5.
FIG. 5.

Effect of local distortion on RMSDs of the five correction methods, L1, L2, GP, MLS, and MLSILS. Thin continuous lines refer to MLS. Thick continuous lines refer to MLSILS. Thin dotted lines refer to GP. Thin dash-dotted lines refer to L1. Thick dash-dotted lines refer to L2.

Image of FIG. 6.
FIG. 6.

Effect of noise on RMSDs of the five correction methods, L1, L2, GP, MLS, and MLSILS. Thin continuous lines refer to MLS. Thick continuous lines refer to MLSILS. Thin dotted lines refer to GP. Thin dash-dotted lines refer to L1. Thick dash-dotted lines refer to L2.

Image of FIG. 7.
FIG. 7.

(a) A distorted image of calibration grid. (b) A distorted image of dry vertebrae. (c) The corresponding image corrected by MLSILS.

Tables

Generic image for table
TABLE I.

Functions and parameters used to generate distortions.

Generic image for table
TABLE II.

RMSDs and ’s of different methods, computed at control points and intermediate points of simulated images.

Generic image for table
TABLE III.

RMSDs and ’s of different methods, computed at control points, intermediate points, and all feature points of simulated images.

Generic image for table
TABLE IV.

RMSDs and ’s of different methods, computed at control points, intermediate points, and all feature points of a real XRII image.

Loading

Article metrics loading...

/content/aapm/journal/medphys/34/11/10.1118/1.2791037
2007-10-16
2014-04-21
Loading

Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: A method based on moving least squares for XRII image distortion correction
http://aip.metastore.ingenta.com/content/aapm/journal/medphys/34/11/10.1118/1.2791037
10.1118/1.2791037
SEARCH_EXPAND_ITEM