Introduction
The Automation Lab focuses mainly on the following AI-based vision processing tasks. In particular, it develops technologies directly related to smart cars and industrial automation, and aims for industry-academia collaboration and commercialization.
1. Vision SoC:
We conduct research on SoC that collect and process images from cameras, sensors, etc., and aim to converge ultra-low power, high-performance hardware and embedded vision algorithms.
2. Automotive ECU Platform and Network:
We develop an electronic control unit and study vehicle network structures.
3. Motion Control:
We study motion control techniques for robots, vehicles, and machine systems, and are implementing feedback control algorithms, stability and responsiveness analysis, and real-time control.
4. Intelligent Vision SW:
We develop environmental recognition technologies such as machine learning-based image processing and object recognition/segmentation, conducting optimization for real-time applications.
Exhibit Item 01
Vehicle Detection & Tracking:
Detect vehicles, keep track of them, draw trajectories, and count the number of them
Speed Estimation:
Estimate the speed of the vehicles using coordinates of detected vehicle bounding boxes and camera parameters
Lane Detection:
Detect lane boundaries from the background of traffic scenery