An evaluation on the effective processing time of Raspberry Pi-3B-based embedded systems for determining forged images

The paper presents the design and processing-time evaluation of an embedded system with the purpose for forensic image detection based on the multi-resolution based algorithm. After the success of the proposed method we then successfully implement on the newest version of the hardware Raspberry Pi-3B which is currently a low cost and friendly environment for variously specific-research applications. The processing times for identifying fake images are compared through different platforms and CPUs of the investigated systems. The results have shown that the processing times depends not only on the dimension and the quality of fake images but also the platform and software facility of the evaluated systems (i.e. on PC-based simulations and Raspberry Pi-based embedded systems).