Poster + Paper
17 June 2024 Single-shot AI-based point spread function method for an active camera alignment
Jiri Thoma, Josef Cupal, Aimira Baitieva, Jakub Cerny, Jan Grundel
Author Affiliations +
Conference Poster
Abstract
In current camera sensor-objective alignment approaches for an automotive industry, slow and iterative scan methods are widely used with a cycle time not fully compatible with contemporary production lines. New method is proposed to facilitate immediate active alignment of the sensor comprising a focal plane array and an objective. The method is based on a single-shot sensor data read-out with unknown absolute and correct alignment positions in all six degrees of freedom. First three unknown alignment values can be calculated based on a classical centroid approach, while the remaining three are calculated using an AI model. This immediately leads to massive production line simplification and cost reduction, so important for the automotive industry.
(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.
Jiri Thoma, Josef Cupal, Aimira Baitieva, Jakub Cerny, and Jan Grundel "Single-shot AI-based point spread function method for an active camera alignment", Proc. SPIE 13024, Optical Instrument Science, Technology, and Applications III, 130240O (17 June 2024); https://doi.org/10.1117/12.3029509
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Cameras

Sensors

Active alignment

Optical alignment

Collimators

Artificial intelligence

LIDAR

Back to Top