Marker Detection

The algorithm performs detection of calibration markers.

Input

Takes a single 2D image set showing a camera calibration target as input. Only 8-bit grayscale and RGB images are supported.

Output

Tracking stream with camera poses for each image in case camera calibration was attached to input images.

Description

Several typical calibration boards are supported. For their overview and parameters please refer to marker configuration.

If checked, the Automatically detect board border checkbox indicates that if not all markers were detected during the first time, the marker detection will run the second time with a cropped image that has a center in a bounding box of detected markers and with double the size of the longer side of the bounding box. Note, this is implemented only for some marker types.

If known, the intrinsic calibration (see camera calibration data component) of the camera used to capture the input images can be set using a camera calibration setter algorithm. If set, this will allow to visualize not only detections but also the reprojections of the marker corners.

When the calibration board parameters and (optionally) camera calibration are set, the Detect button will run the detection of the markers. The checkboxes Detections and Reprojections can be toggled to visualize the corresponding points on top of the images. Note, that toggling of Reprojections will have no effect if the calibration was not set.