Bibliographic Details
| Title: |
A parallel texture-based region-growing algorithm implemented in OpenMP: A parallel texture-based region-growing...: C. Couder-Castañeda et al. |
| Authors: |
Couder-Castañeda, Carlos1 (AUTHOR), Orozco-del-Castillo, Mauricio2 (AUTHOR) mauricio.orozco@itmerida.edu.mx, Padilla-Perez, Diego1 (AUTHOR), Medina, Isaac1 (AUTHOR) |
| Source: |
Scientific Reports. 2/14/2025, Vol. 15 Issue 1, p1-16. 16p. |
| Subject Terms: |
*PARALLEL programming, *THREE-dimensional imaging, *SEQUENCE analysis, *FORTRAN, *DATA extraction |
| Abstract: |
We introduce an innovative OpenMP design and implementation of a search-growing algorithm for fine-grained lock-based parallelization, which enhances automatic volume extraction in seismic data represented by 3D voxel pictures to detect salt bodies and assess the voxel texture. The first algorithm was developed in MATLAB, subsequently translated to Fortran, and ultimately enhanced with OpenMP pragmas to facilitate the parallel version. In the parallel version, synchronization techniques are essential to enforce limits on the sequence of voxel analysis conducted by the threads. To do this, we employed OpenMP locks to ensure that only a single thread can process one voxel, hence maintaining the coherence of the data. Minimizing crucial regions in parallel programming is advisable as they might hinder scalability and performance. In this application, locks do not induce latency since not all threads attempt to access the same voxel, indicating that not all threads traverse the same critical region. Our design involves securing each voxel that may adapt to the 3D image, enhancing performance and markedly decreasing execution time. Performance trials were conducted on a dual Xeon system featuring twenty physical cores, with Hyperthreading (HT) both disabled and enabled. In this investigation, HT improved slightly the performance. [ABSTRACT FROM AUTHOR] |
|
Copyright of Scientific Reports is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) |
| Database: |
Academic Search Complete |
|
Full text is not displayed to guests. |
Login for full access.
|
