TY - GEN
T1 - Parallelization of 3D-LBP Feature Extraction Using MPI Programming
AU - Prabhu, Vishwas
AU - Marate, Nithanth
AU - Ashwath, Rao B.
AU - Gopalakrishna Kini, N.
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2026.
PY - 2026
Y1 - 2026
N2 - It is difficult to identify criminals using biometric verification; high precision is needed to improve public safety and minimize misidentification. The 3D-Local Binary Pattern (3D-LBP) feature descriptor is widely employed in biometric techniques such as fingerprint identification, iris recognition, and facial recognition. A straightforward and effective technique for characterizing local picture content with significant discriminative power is 3D-LBP. In order to extract intricate spatial and temporal information from facial photos, this study uses 3D-LBP to investigate pixel interactions in a three-dimensional framework. The Message Passing Interface (MPI) is used for distributed computing to increase the performance of facial recognition. MPI makes face recognition and feature extraction real-time possible, making it possible to distribute computing work among several processors. Using MPI results in noticeable speedups of around five times, showcasing important developments in face recognition technology that are necessary for contemporary security systems. This make the study to enhance the utilization of the computing power of the CPU so that it can be used in the security systems which now are in high demand.
AB - It is difficult to identify criminals using biometric verification; high precision is needed to improve public safety and minimize misidentification. The 3D-Local Binary Pattern (3D-LBP) feature descriptor is widely employed in biometric techniques such as fingerprint identification, iris recognition, and facial recognition. A straightforward and effective technique for characterizing local picture content with significant discriminative power is 3D-LBP. In order to extract intricate spatial and temporal information from facial photos, this study uses 3D-LBP to investigate pixel interactions in a three-dimensional framework. The Message Passing Interface (MPI) is used for distributed computing to increase the performance of facial recognition. MPI makes face recognition and feature extraction real-time possible, making it possible to distribute computing work among several processors. Using MPI results in noticeable speedups of around five times, showcasing important developments in face recognition technology that are necessary for contemporary security systems. This make the study to enhance the utilization of the computing power of the CPU so that it can be used in the security systems which now are in high demand.
UR - https://www.scopus.com/pages/publications/105029560578
UR - https://www.scopus.com/pages/publications/105029560578#tab=citedBy
U2 - 10.1007/978-981-95-2878-3_1
DO - 10.1007/978-981-95-2878-3_1
M3 - Conference contribution
AN - SCOPUS:105029560578
SN - 9789819528776
T3 - Lecture Notes in Networks and Systems
SP - 1
EP - 12
BT - Computing and Machine Learning - Proceedings of CML 2025
A2 - Bansal, Jagdish Chand
A2 - Borah, Samarjeet
A2 - Hussain, Shahid
A2 - Salhi, Said
PB - Springer Science and Business Media Deutschland GmbH
T2 - 2nd International Conference on Computing and Machine Learning, CML 2025
Y2 - 22 March 2025 through 23 March 2025
ER -