Louisa Crawford
27 Aug 2024 20:32
NVIDIA has introduced a streamlined camera calibration process to improve the accuracy of AI-based multi-camera tracking applications.
NVIDIA has revealed advances in camera calibration aimed at increasing the accuracy and efficiency of AI-based multi-camera tracking applications. According to the NVIDIA Technical Blog, the development is part of an ongoing effort to streamline the process within the company’s Metropolis framework.
Camera Calibration
Camera calibration is essential to convert the 2D camera view into real-world coordinates to enable accurate object tracking and localization. This process involves determining certain camera parameters, which are divided into extrinsic and intrinsic categories. Extrinsic parameters define the position and orientation of the camera relative to the world coordinate system, while intrinsic parameters map camera coordinates to pixel coordinates.
Correction in multi-camera tracking
NVIDIA Metropolis uses calibrated cameras as sensors to enhance spatial-temporal analysis in multi-camera AI workflows. Proper camera calibration is essential to accurately locate objects within a coordinate system and facilitate key functions such as location services, activity correlation across multiple cameras, and distance-based metric calculations.
For example, in a retail store, a calibrated camera can locate customers on a floor plan. In a warehouse, multiple calibrated cameras can track people moving across different areas, ensuring seamless monitoring. Calibrated cameras also enable accurate distance calculations, as they eliminate variability due to pixel domain mismatch.
Metropolis Camera Calibration Toolkit
NVIDIA’s Metropolis Camera Calibration Toolkit simplifies the calibration process by providing tools for project setup, camera import, and reference point selection. It supports three calibration modes: Cartesian calibration, multi-camera tracking, and image. The toolkit ensures that cameras are accurately calibrated, generating formatted files compatible with other Metropolis services.
Users can start by importing a project with the provided assets or creating one from scratch. The calibration process involves selecting reference points that are visible in both the camera image and the floor plan, and creating a transformation matrix to map the camera trajectory to the floor plan. The toolkit also provides add-ons for regions of interest (ROIs) and tripwires to enhance its usability for a variety of applications.
Auto correction for synthetic cameras
NVIDIA Metropolis also supports synthetic data via the NVIDIA Omniverse platform. omni.replicator.agent.camera_calibration The extension automates the calibration of synthetic cameras, eliminating the need for manual reference point selection. The tool outputs the required mapping with a single click, making it easier to integrate synthetic video data into Metropolis workflows.
The automatic calibration process involves creating a top-view camera and automatically selecting reference points to calibrate other cameras. This extension computes the intrinsic and extrinsic matrices of the camera, the projection matrix, and the correspondence between the camera view and the plan map, exporting them as JSON files for seamless integration.
conclusion
Camera calibration is a key step in enhancing the capabilities of the NVIDIA Metropolis application, enabling accurate object positioning and correlation across multiple cameras. This advancement paves the way for large-scale real-time location services and other intelligent video analytics applications.
For more information and technical support, visit the NVIDIA Developer Forums.
Image source: Shutterstock