transmote speaks…

FLARManager API Documentation

The complete FLARManager ASDoc reference can be found here:
http://transmote.com/flar/reference

FLARManager config file

The preferred method for initializing FLARManager is to pass the path to an external XML configuration file into the FLARManager constructor:

var flarManager:FLARManager = new FLARManager("flarConfig.xml");

In the FLARManager distribution, flarConfig.xml is located in /resources/flar/, so the whole path relative to the bin-debug folder is ../resources/flar/flarConfig.xml, which yields:
new FLARManager("../resources/flar/flarConfig.xml");

The configuration file can be called whatever you like, as long as you format it according to the following specifications. There are four top-level nodes, and each node accepts a number of attributes. The <patterns> node also accepts a list of child nodes. Attributes for these nodes are listed below, along with their types and default values. Required nodes and attributes are marked as such; if a non-required attribute is omitted, its default value will be used.

<flarSourceSettings>

[attributes]

• sourceWidth:Number=640
  Width at which to capture video. This is the value passed to Camera.setMode() in FLARCameraSource.as.
   
• sourceHeight:Number=480
  Height at which to capture video. This is the value passed to Camera.setMode() in FLARCameraSource.as.
   
• displayWidth:Number=640
  Width at which to display video.
   
• displayHeight:Number=480
  Height at which to display video.
   
• framerate:Number=30
  Framrerate at which to capture video. This is the value passed to Camera.setMode() in FLARCameraSource.as.
   
• downsampleRatio:Number=0.5
  The amount to downsample camera input. The captured video is scaled down by this value before being sent to FLARToolkit for analysis. FLARToolkit runs faster with more downsampling, but also has more difficulty recognizing marker patterns. A value of 1.0 results in no downsampling; a value of 0.5 (the default) downsamples the camera input by half.
   
• loaderPath:String=""
  If this value is set, FLARManager will create a FLARLoaderSource instead of the default FLARCameraSource. The file located at this path will be used by the FLARLoaderSource.
   
• useProxy:Boolean=false
  If true, FLARManager will create a FLARProxy instead of the default FLARCameraSource.
   

[children]
none.

<flarManagerSettings>

[attributes]

• mirrorDisplay:Boolean=true
  If true, video and detected marker data are flipped horizontally. (NOTE: As of v0.6, transformation matrices are flipped correctly, and developers no longer need to flip the DisplayObjectContainer into which the transformed objects are drawn.)
   
• smoothing:int=3
  Apply a smoothing algorithm to transformation matrices generated by FLARToolkit. Smoothing is equal to the number of frames over which FLARManager will average transformation matrices; the larger the number, the smoother the animation, and the slower the response time between marker position/orientation changes. A value of 0 turns smoothing off.
   
• adaptiveThresholding:Boolean=true

  NOTE: this attribute is deprecated and removed in v0.6. Use thresholdAdapter instead.
  Adaptive thresholding can result in better marker detection across a range of illumination. This is desirable for applications with low lighting, or in which the developer has little control over lighting conditions, such as with web applications.
   

• adaptiveThresholdingSpeed:Number=0.3

  NOTE: this attribute is deprecated and removed in v0.6. Use thresholdAdapter instead.
  The speed at which the threshold changes during adaptive thresholding. Larger values may increase the speed at which the markers in uneven illumination are detected, but may also result in instability in marker detection.
   

• adaptiveThresholdingBias:Number=-0.1

  NOTE: this attribute is deprecated and removed in v0.6. Use thresholdAdapter instead.
  The direction toward which adaptive thresholding trends. Lower thresholds favor environments with too little light; higher thresholds favor environments with too much light.
   

• sampleBlurring:int=1
  The amount of blur applied to the source image before sending to FLARToolkit for marker detection. Higher values increase framerate, but reduce detection accuracy.
   
• markerUpdateThreshold:Number=80
  If a detected marker is within this distance (pixels) from an active marker, FLARManager considers the detected marker to be an update of the active marker. Else, the detected marker is a new marker. Increase this value to accommodate faster-moving markers; decrease it to accommodate more markers on-screen at once.
   
• markerRemovalDelay:int=1
  Number of frames after removal that a marker will persist before dispatching a MARKER_REMOVED event.
   
• minimumLabelSize:int=100
  Provides direct access to FLARLabeling_BitmapData.minimumLabelSize, which is the minimum size (width*height) a dark area of the source image must be in order to become a candidate for marker outline detection. Higher values result in faster performance, but poorer marker detection at smaller sizes (as they appear on-screen).
   

[children]

• smoother
  IFLARMatrixSmoother to use to apply smoothing to transformation matrices generated by FLARToolkit. This defaults to FLARMatrixSmoother_Average, but developers can implement their own smoothing algorithms. See Inside FLARManager: Customization for more information.
   
• [attributes]
  • name:String='FLARMatrixSmoother_Average'
    The name of the smoother class to use. If the class resides somewhere other than com.transmote.flar.utils.smoother, a fully-qualified class name must be used.
     
• thresholdAdapter:String='DrunkWalkThresholdAdapter'
  Adaptive thresholding can result in better marker detection across a range of illumination. This is desirable for applications with low lighting, or in which the developer has little control over lighting conditions, such as with web applications.
  The default threshold adapter is DrunkWalkThresholdAdapter, but developers can implement their own adaptive thresholding algorithms. See Inside FLARManager: Customization for more information.
   
• [attributes]
  • name:String='DrunkWalkThresholdAdapter'
    The name of the threshold adapter class to use. If the class resides somewhere other than com.transmote.flar.utils.threshold, a fully-qualified class name must be used.
     
  • speed:Number=0.3
    The speed at which the threshold changes during adaptive thresholding. Larger values may increase the speed at which the markers in uneven illumination are detected, but may also result in instability in marker detection.
     
  • bias:Number=-0.1
    The direction toward which adaptive thresholding trends. Lower thresholds favor environments with too little light; higher thresholds favor environments with too much light.
     

<cameraParamsFile>

–REQUIRED–
[attributes]

• path:String="" –REQUIRED–
  The path to the FLARToolkit camera parameters file (e.g. FLARCameraParams.dat or camera_para.dat). FLARToolkit requires this file to initialize and run.
   

[children]
none.

<patterns>

–REQUIRED–
[attributes]

• resolution:int=0 –REQUIRED–
  Resolution (width/height) of marker pattern files. All patterns in a single application must share the same resolution; this is mandated by FLARToolkit. For more information about pattern resolution, see here.
   
• patternToBorderRatio:Number=50
  Out of the entire width/height of a marker, the amount that the pattern occupies relative to the amount the border occupies. Value is expressed as a percentage (0-100). For example, a value of 50 indicates that the width of the pattern area is equal to the total width (on either side of the pattern) of the border. NOTE: this is not yet implemented by FLARToolkit for multiple markers, and is therefore not yet implemented for FLARManager.
   
• minConfidence:Number=0.5
  ‘Confidence’ is a value (from 0.0 to 1.0) assigned by FLARToolkit to each detected marker, that describes the algorithm’s perceived accuracy of the pattern match. This value sets the minimum confidence required to signal a recognized marker. Lower values will be more likely to detect both real markers and false positives; higher values will be less likely to detect both real markers and false positives.
   
• size:Number=80
  The width of a marker (in pixels) on-screen at which the scale of its transformation matrix is 1.0. This can be used to adjust the size that a mode will appear on-screen. If a model is appearing too large, decrease this number, and vice-versa.
   

[children]

• <pattern>
  Each <pattern> node represents a pattern that FLARToolkit will attempt to detect. FLARToolkit can look for an unlimited number of patterns, but as it attempts to match each detected marker against all possible patterns, more patterns mean slower pattern matching (and ultimately, slower overall performance). Developers should keep this list as short as possible for each application. Each <pattern> node has one attribute, path, that describes the path to the associated pattern file (relative to the published SWF).