Config file schema definition for the tag-grid-scanner.
The configuration of the camera.
The index of the camera.
This is system dependent and different devices might be associated with the same index during the runtime of the program. The index and how it is changing is highly system-dependent, but should be safely usable if there is only one camera connected to the system.
Value must be greater or equal to 0
The size of the requested camera video stream.
The given size must match one of the supported resolutions of the camera. The result of querying an unsupported size is undefined.
Must contain a minimum of 2 items
Must contain a maximum of 2 items
A positive integer, i.e. it an integer > 0.
Value must be greater or equal to 1
The number of frames per second of the requested camera video stream.
The given number must match one of the supported number of fps of the camera. The result of querying an unsupported fps is undefined.
Value must be greater or equal to 1
The 4-digit code of the requested camera video stream.
A camera might support different stream formats. Selecting a specific format might affect image quality and performance. The most common ones for webcams are 'MJPG' and 'H264'. See also the list of known FOURCCs.
Must be at least 4 characters long
Must be at most 4 characters long
Calibration data to remove distortion from the image.
Camera lenses are not perfect. Instead, they always introduce some kind of distortion to the image captured by the camera sensor. If the cameras corrects the distortion internally, the calibration data can be ignored, but usually, even the internal correction is not good enough to have perfectly map straight lines in the scene to perfectly straight lines in the camera image. The objects in this section can be generated using the calibrate command of the tag-grid-scanner. See also OpenCV's documentation on camera calibration.
A 3x3 matrix of floating point numbers.
Must contain a minimum of 3 items
Must contain a maximum of 3 items
Must contain a minimum of 3 items
Must contain a maximum of 3 items
Must contain a minimum of 5 items
Must contain a maximum of 5 items
The name of a video file, image or camera device node.
Use this instead of the camera index. It allows to read image data from video files, images or from camera device nodes on certain operating systems such as Linux. The latter is very useful to make sure the correct camera is used in multi-camera systems, since the camera index may change arbitrarily. It is common practice in Linux to wait for a certain camera via UDEV and create a well-defined symlink to the camera device node that can be used reliably here.
Scale factor applied to the camera image.
If the camera only provides a fixed resolution or lower than maximum resolution just crop the image instead of scaling it, this scale factor can be used. Lower values will improve the speed of the computation. However, values too low will result in tag detection errors.
A non-negative floating point numner, i.e. a floating point number >= 0.0.
Value must be greater or equal to 0.0
A 2D vector of non-negative floating point numbers.
Must contain a minimum of 2 items
Must contain a maximum of 2 items
A non-negative floating point numner, i.e. a floating point number >= 0.0.
Value must be greater or equal to 0.0
Angle of rotation applied to the camera image.
Whether to flip the camera image vertically.
Flipping the image needs to be enabled if the camera is seeing the ROI via a mirror, which results in all tags being flipped, too. Flipping the grid (but not the tags) can be reverted in the notify section.
Whether to flip the camera image horizontally.
Flipping the image needs to be enabled if the camera is seeing the ROI via a mirror, which results in all tags being flipped, too. Flipping the grid (but not the tags) can be reverted in the notify section.
Blend the current image with previous images.
The current image will be blended with the previous image using the smooth value as the interpolation parameter. At 0.0, the current image will be used exclusively, at 1.0, the previous will be used exclusively (which makes no sense). Values of 0.5 to 0.75 worked reasonably well for reducing noise in the input images due to insufficient lighting. Values closer to one will increase detection latency. It is important to consider this option in the light of the camera's fps, since a constant smooth value will result in different latencies for different fps.
Value must be greater or equal to 0.0 and lesser or equal to 1.0
The configuration of the tag grid.
The size of the tag grid.
How many tags can be placed in each direction?
Must contain a minimum of 2 items
Must contain a maximum of 2 items
A positive integer, i.e. it an integer > 0.
Value must be greater or equal to 1
The size of each tag.
How many pixels does each tag have in each direction?
Must contain a minimum of 2 items
Must contain a maximum of 2 items
A positive integer, i.e. it an integer > 0.
Value must be greater or equal to 1
The region of interest for the scanner to look for tags.
Provide the vertices of a quadrilateral containing the tag grid in the image. The vertices are relative to the size of the image, i.e. each value should be in [0, 1].
Must contain a minimum of 4 items
Must contain a maximum of 4 items
Must contain a minimum of 2 items
Must contain a maximum of 2 items
Size of the gaps in between tags.
The gap size is relative to the image size, i.e. each value should be in [0, 1].
Must contain a minimum of 2 items
Must contain a maximum of 2 items
A floating point number in [0, 1].
Value must be greater or equal to 0.0 and lesser or equal to 1.0
Keep this fraction of the center part of each tile's pixel.
In most cases, the tag pixels are not perfectly aligned with the grid in the image. This is usually due to imperfection in the camera lens, calibration or selection of the region of interest. To attenuate this problem, each tag's pixels can be cropped such that only their center part is used for identifying the pixels color.
A floating point number in [0, 1].
Value must be greater or equal to 0.0 and lesser or equal to 1.0
A 2D vector of floating point numbers in [0, 1].
Must contain a minimum of 2 items
Must contain a maximum of 2 items
A floating point number in [0, 1].
Value must be greater or equal to 0.0 and lesser or equal to 1.0
Define how the scanner should notify about detected tags.
If the conditions in the "If" tab are respected, then the conditions in the "Then" tab should be respected. Otherwise, the conditions in the "Else" tab should be respected.
true If the conditions in the "If" tab are respected, then the conditions in the "Then" tab should be respected. Otherwise, the conditions in the "Else" tab should be respected.
true The JSON template to which the tag data will be added.
When the scanner outputs a notification, it will use this object as a template and assign the tag data to the JSON pointer defined by the assignTo option.
JSON pointer into the template option for assigning the tag data during notification.
When the scanner outputs a notification, it will copy the object provided via the template option and assign the tag data to the property defined by this JSON pointer. See also the definition of the JSON pointer format.
Rotate the grid by the given number of degrees.
Flip the grid vertically.
Flip the grid horizontally.
Output the notification to the standard output stream.
Output the notification to the standard error stream.
Send the notification to a remote host via HTTP.
Repeat the notification even though the detected tags did not changed.
Interval at which notification are repeated in seconds, when repeat is enabled.
Value must be greater or equal to 1