FPGA Implementation of Sketched LiDAR for a 192 x 128 SPAD Image Sensor
This breakthrough slashes the data transfer barrier holding back next-gen sensors.
Researchers have developed a new FPGA-based compression algorithm for high-resolution SPAD LiDAR sensors, which can generate data at tens of gigabytes per second. Their hardware implementation achieves a massive 512x compression ratio compared to conventional histogram outputs, enabling real-time, high-fidelity depth reconstruction directly from sensor timestamps. This directly tackles the critical data transfer bottleneck, allowing these powerful sensors to scale as pixel counts increase for future applications like autonomous vehicles.
Why It Matters
It unlocks the potential of ultra-high-resolution LiDAR for real-time use in robotics and self-driving cars by solving a fundamental data throughput problem.