Continental Automotive Gmbh

Continental Autonomous Mobility Germany Gmbh

Alexander Kadyrov

Rui Guerreiro

Saharsh Bishnoi

Robert Gillott

Nauman Javed

Rodolfo Cignoli

Abhinav Pandey

A method for determining extrinsic camera parameters includes: starting camera movement, capturing a first raw image with parallel first and second patches at a first point of time, and a second raw image with parallel third and fourth patches at a second point of time. A distance between the first and second patches and between the third and fourth patches is the same. A reference position A of a first patch image feature, a reference position C of a second patch image feature, an offset position B of the feature of the first patch in the third patch, and an offset position D of the feature of the second patch in the fourth patch are detected. An epipolar model is applied based on the positions A-D and a distance travelled by the camera between the first and second time points. Extrinsic camera parameters are determined from the epipolar model.

System and method for calibrating a camera in a vehicle

A method for detecting contaminations on a lens of a vehicle camera. Transition values are assigned to pixels of image frames. Three-dimensional coordinates are assigned to the pixels by using a simplified model of the scene. Using the assigned coordinates and movement data, trajectories are assigned to the pixels. An expectation value is derived from a transition value of a starting pixel of a trajectory and a satisfied transition value is derived from transition values along the trajectory. The expectation value and the satisfied transition value are accumulated over time, and a cleanliness value is derived from a ratio of the accumulated satisfied transition value to the accumulated expectation value.

Detection of lens contamination using expected edge trajectories

Growing community of inventors

Eastbourne, United Kingdom

Alexander Kadyrov