Real-Time Information Fusion System Implementation Based on ARM-Based FPGA

In this study, an information fusion system displayed on a transparent display by considering the relationships among display, background exhibit, and user’s gaze direction. We used ARM-based field-programmable gate array (FPGA) to perform virtual–real of as well evaluated execution speed. The FPGA Intel® RealsenseTM D435i depth cameras capture color images observer exhibit. image data was received ARM side fed for real-time object detection. accelerated computation convolution neural networks recognize observers exhibits. addition, module performed developed rapid registration between images. calculated size position according pixel coordinates values human eye A personal computer with GPU RTX2060 in ~47 ms, whereas accomplished it 25 ms. Thus, speed 1.8 times faster than computer.