Table 2
Comparison between memory usage of A-stacking and A-projection forward calculations based on (23)and (24), respectively, and using K = 9121 unique baselines.
| Method | Np = 64 | Np = 1024 | ||
|
|
||||
| A-stack, nB = 10 | 2.2 MB | 177.5 MB | ||
| A-stack, nB = 20 | 4.2 MB | 338.9 MB | ||
| A-stack, nB = 40 | 8.3 MB | 661.7 MB | ||
| A-stack, nB = 96 | 19.6 MB | 1565.5 MB | ||
| A-stack, nB = 192 | 38.9 MB | 3114.9 MB | ||
| A-proj, NgA = 25 | 87.0 MB | 103.0 MB | ||
| A-proj, NgA = 50 | 348.1 MB | 364.1 MB | ||
| A-proj, NgA = 100 | 1391.8 MB | 1407.8 MB | ||
Notes. Results are shown for small and moderate size images, and for various levels of accuracy in modeling BD-DD gains (number of terms nB for A-stacking, and convolution kernel size NgA for A-projection). Double precision computation is assumed.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.