개요

How can massive amounts of data from camera streams be efficiently captured without storage and processing costs skyrocketing? The hands-on video on AUTERA real-time data compression addresses this challenge.

In modern data acquisition campaigns involving multiple vehicles and high-resolution sensors, raw data quickly consumes terabytes of storage and leads to long upload and processing times. AUTERA provides an efficient, practical solution here. It uses GPU-accelerated compression on the AUTERA AutoBox, allowing video data to be compressed live during recording. Using RTMaps and GStreamer integration, the workflow demonstrates how H.264 compression can be seamlessly integrated into an existing data acquisition pipeline with minimal configuration effort.

A realistic setup for autonomous driving featuring 11 cameras and 8 lidar sensors demonstrates the system’s capabilities. Data volume can be reduced to just a few percent of the raw data size while maintaining visually lossless quality. This significantly reduces the required SSD storage, accelerates data transfer, and improves the overall efficiency of the workflow.

This video is of particular interest to engineers and managers involved in ADAS/AD data acquisition, regardless of their industry. In addition to the automotive sector, companies in the agricultural, off-highway, and defense industries can also benefit from it. It offers a scalable, cost-effective approach to handle high-bandwidth sensor data without compromising its usability for development, testing, and validation.