High-speed FPGA-based flow detection for microfluidic Lab-on-Chip

Machine vision algorithms provide significant benefits for Lab-on-Chip (LoC) systems by automating the experimental process. This paper presents an FPGA-based machine vision flow detection implementation for microfluidic Lab-on-Chip (LoC) experiments. We propose and implement a novel architecture that exploits modern FPGA parallelism capabilities and makes efficient use of device resources to achieve real-time data collection in megapixel resolutions, at rates exceeding 30000 frames per second.