Asynchronous Processing for Latent Fingerprint Identification on Heterogeneous CPU-GPU Systems

Abstract

Latent fingerprint identification is one of the most essential identification procedures in criminal investigations. Addressing this task is challenging as (i) it requires analyzing massive databases in reasonable periods and (ii) it is commonly solved by combining different methods with very complex datadependencies, which make fully exploiting heterogeneous CPU-GPU systems very complex. Most efforts in this context focus on improving the accuracy of the approaches and neglect reducing the processing time. Indeed, the most accurate approach was designed for one single thread. This work introduces ALFI (Asynchronous processing for Latent Fingerprint Identification), the fastest methodology for latent fingerprint identification maintaining high accuracy. ALFI fully exploits all the resources of CPU-GPU systems using asynchronous processing and fine-coarse parallelism for analyzing massive databases. Our approach reduces idle times in processing and fully exploits the inherent parallelism of comparing latent fingerprints to fingerprint impressions. We analyzed the performance of ALFI on Linux and Windows operating systems using the well-known NIST/FVC databases. Experimental results reveal that ALFI is in average 22x faster than the state-of-the-art algorithm, reaching a value of 44.7x for the best-studied case.

Fast and accurate phase processing in off-axis digital holography combining adaptive spatial filtering and an embedded GPU platform

“…In designing CPU and GPU acceleration strategies, tasks are assigned based on the computational characteristics of the CPU and GPU [38][39][40]. The CPU is suitable for a series of control-type tasks, particularly those requiring low latency and high performance, such as data transfer and image display processes.…”

“…The CPU and GPU performances can be optimized by disabling the power-saving functions of the CPU, adjusting the GPU clock to the maximum frequency, and closing all superfluous programs. Due to the independent CPU and GPU memories on computers, there are many migrations between them; the CPU and GPU of the Jetson Nano Developer Kit share physically unified memory, eliminating the tedious transfer process [35][36][37][38][39][40].…”

Face Recognition on a Smart Image Sensor Using Local Gradients

“…The technological advances in high-performance computing have enabled the development of fast and highly accurate biometric systems. Most frequently, this performance is achieved using power-hungry processors and graphics processing units (GPUs) [ 42 , 43 ]. While this cost in power and space may not be important in big data applications that require high precision, it is normally not acceptable in mobile or portable biometric systems [ 11 ], which require compact, power-efficient electronics.…”