Imatest Launches Camera Geometric Calibration Validation service for ADAS and Autonomous Vehicle
Service uses upcoming IEEE P2020 geometric calibration standard as a reference
Imatest, LLC unveiled their new Camera Geometric Calibration Validation service to provide objective metrics for calibration quality. Calibration is necessary to provide correspondences between image and world points when building an automotive camera system, providing a way to determine that a calibration describes the geometric camera behavior at real-world distances. Another benefit is to be able to compare calibration methods, models, and/or parameterizations, checking coordinate system consistency. Independent testing can help reveal internal model discrepancies that would not otherwise be discovered.
Imatest’s new facilities in Boulder, Colorado includes a permanent test capability for validation at application distances up to 20 meters. The test setup and method will be consistent with the upcoming IEEE P2020 Automotive Imaging standard.
The Imatest method produces 64 target pair combinations for analysis. Between-target distance error is used to characterize the quality of a given calibration. The camera is rotated to many positions to evaluate through the image field, and each image produces a population of distance error values for analysis.
Original release: PR Newswire
