Automotive & Industry 4.0

Automated driving: video systems development

Conference Hall

21st November, 12:00-12:30

In today's world we find ourselves inside a transitional period: moving from human-based driving towards fully automated driving, with current state of the art being partly-autonomous vehicles. It is widely considered that the goal of autonomous driving can only be achieved by using multiple sensors: video, radar/lidar, ultrasonic, etc. Amongst these, the video sensor (onboard camera) is one of the most important ones, and is also the aim of this presentation. It can be used to detect the drivable space, objects on it (e.g. cars, pedestrians), lane markings, road signs and other important information found on the road. In Cluj, we contribute to developing video-based systems, using both computer vision algorithms, as well as deep learning/machine learning technique, which are fused together to form a robust system.

Lucian Cristea


Strive for excellence.
As a SW developer at Bosch Engineering Center Cluj, and as a Master's student at UTCN Computer Science department, it is my belief that in order to achieve true greatness, one must explore all branches of the technical field. At Bosch, I work on both classical model-based image-processing techniques and on deep learning / machine learning algorithms. Through a combination of the two, we build a system capable of performing key human driver tasks autonomously, such as emergency braking or lane keeping, while being both performant and robust. My goal is to actively contribute to this autonomous-vehicle field, and in the process of doing so, learn from it.