Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Developing a platform for multi-camera SLAM (collaboration with Sony)
Develop a hardware synchronized system of stereo cameras and IMU to simultaneously process images from multiple cameras and IMU measurements on-board in real time.
Stereo cameras are used in SLAM pipelines to provide robustness and accurate scale even under challenging motions such as pure rotations. Integrating IMU measurements with stereo camera allows for accurate tracking performance in SLAM, especially when the platform undergoes fast motion. There have been successful commercial applications of visual SLAM with stereo sensors. For example, Skydio r1 drone uses 12 cameras for navigation to ensure robustness. In this project, we will develop a hardware synchronized system of stereo cameras and IMU to process the camera images from multiple stereo pairs and IMU measurements on-board in real time.
Stereo cameras are used in SLAM pipelines to provide robustness and accurate scale even under challenging motions such as pure rotations. Integrating IMU measurements with stereo camera allows for accurate tracking performance in SLAM, especially when the platform undergoes fast motion. There have been successful commercial applications of visual SLAM with stereo sensors. For example, Skydio r1 drone uses 12 cameras for navigation to ensure robustness. In this project, we will develop a hardware synchronized system of stereo cameras and IMU to process the camera images from multiple stereo pairs and IMU measurements on-board in real time.
The goal of this project is to develop a pipeline that can simultaneously record images from multiple stereo cameras and IMU and perform some basic image processing on board in real time.
Applicants should have experience with designing real-time systems, ROS and a strong background in C/C++. Experience with designing drivers for embedded platforms is a plus.
The goal of this project is to develop a pipeline that can simultaneously record images from multiple stereo cameras and IMU and perform some basic image processing on board in real time. Applicants should have experience with designing real-time systems, ROS and a strong background in C/C++. Experience with designing drivers for embedded platforms is a plus.