NaturalPoint Tracking Toolkits (Point Cloud and Rigid Body) Users Manual   4. Using the Point Cloud Software The Point Cloud toolkit is a powerful solution for your advanced tracking needs. It provides 3D tracking and calibration across multiple OptiTrack cameras in a compact software toolkit, all with an easy to use API. With the Point Cloud toolkit properly installed and calibrated, your applications will have easy access to robust and precise tracking data. The information provided below will help ensure that you get the best results from your OptiTrack cameras and Point Cloud software. 4.1 Getting Started Before using the Point Cloud software you should make sure to have all of the software and hardware properly installed and licensed as described in Chapter 3. Once the installation is complete you can begin using the Point Cloud software by following these instructions :   Connect 3 or more cameras to the USB port of the computer and arrange them to set up a capture volume as described in Section 3.5 ("Camera Placement").   Run the calibration wizard to calibrate the cameras and test the capture volume.   Save the resulting calibration for later use in your own software. 4.2 Camera Calibration Tool Before the Point Cloud software is able to track markers, it necessary to conduct a camera calibration first. The calibration allows you to fine-tune the OptiTrack camera settings, identifies the physical location of the cameras (extrinsic parameters) and the properties of their lenses (intrinsic parameters). The Calibration tool which guides the process also provides a way to review the results and capture sample data. Before starting the calibration you will need to have your cameras set up as described in Section 3.5 ("Camera Placement"). If any of the cameras are physically moved, it will be necessary to conduct a new calibration.   The following section describes how to use the calibration tool. To start the tool, click the "OptiTrack Point Cloud Calibration Tool" icon on your desktop.     4.2.1 Licensing The first step of the Camera Calibration verifies that a valid license and matching camera are present. It also checks to make sure that 3 or more cameras are connected to the computer. It is not possible to conduct the calibration and track with less than three cameras.     If everything is ready, press the "Begin" button to proceed to the next step.   4.2.2 Camera Preview The Camera Preview step displays the scene contents from the perspective of the cameras. The purpose of this step is to remove all unwanted light picked up by the cameras. This can be done by removing physical items, turning off lights, covering windows, adjusting camera settings or using the blocking tools. Please be sure your ground plane and wand are not in the capture volume during this process.     Camera Views : Moving the mouse over an individual camera view will temporarily switch the camera into greyscale mode and display the actual capture volume from that camera's perspective. A camera view can be zoomed to and from full screen by double clicking in its border area.   Camera Settings The OptiTrack cameras have settings which can be optimized to help reduce unwanted light in the scene. When adjusting the settings it is possible to change them for individual cameras, or all at the same time using the "Camera Settings" dropdown selector.   Threshold : This scrollbar adjusts how bright a marker has to be before it will be identified by the cameras. Using a larger value such as "180" will help reduce the number of unwanted items in the scene. Using too high of a value (over "230") can degrade the precision of the detected markers.   Exposure : This scrollbar adjusts the camera's shutter timer which determines the amount of light a camera lets in during each frame. Using a smaller value such as "55" will help reduce the number of unwanted items detected in the scene.   Illumination : This scrollbar adjusts the brightness of the illumination provided by the OptiTrack camera's on board IR LEDs.   For best performance when using OptiTrack FLEX:V100 cameras, use the IR strobe mode. This mode is activated by setting the illumination value to "15". When strobe mode is active there is no benefit to using an exposure value larger than "100" since the LEDs cease to emit light after that time. To adjust the brightness of the LEDs in strobe mode, modify the exposure setting within the range of "1" through "100" (smaller values will result in less light).   Blocking tools The software can be instructed to ignore detected light (markers, light sources, etc) in the field of view of a camera by using blocking regions. Blocking regions can be added by manually drawing boxes on the camera previews using the mouse or automatically using the "Block Visible Markers" button. To remove all blocking regions press the "Clear Block Regions" button.   Calibration Options The quality of the calibration results can be improved taking more time to complete. Typically the faster settings will be sufficient. Use the "Quality" dropdown selector to choose the desired calibration level.   Once the scene is clear and has all markers either removed or blocked, press the "Next" button to proceed.   4.2.3 Wand Capture The purpose of the Wand Capture step is to collect a set of data from the cameras which will be used to calculate their physical position and lens characteristics. A wand (reflective marker with a handle) should be moved through the capture volume in a three dimensional pattern at a moderate speed while the software records it. As the wand is moved through the scene it will be represented on the screen as a circle with a different color for each camera. When a sample point is taken it will show up on the screen in a darker color.   Special attention should be paid to obtaining good coverage for each camera. The recorded sample points can be used as a visual indicator and guide of wand coverage. Moving the wand in a circular, three dimensional pattern can also help in obtaining good results. Also try to wave the wand away and toward the cameras often to produce as much depth motion as possible.   It is important that the wand is the only detected marker in the scene for every camera. Additional markers can introduce errors and reduce the quality of the calibration. The software will warn the user and attempt to ignore additional markers, however if additional markers are present it is recommended to return to the previous "Camera Preview" step and remove them.     Wand Data When you are ready to start the wand data capture process, press the Start button. As the data is collected the progress bar will be updated to indicate how much time remains.   Synchronization The synchronization indicator provides information about how well the OptiTrack cameras are synchronized. Using hardware camera synchronization can greatly improve the performance of the calibration and 3D tracking. Please see Section 3.5 ("Camera Synchronization") for more information.   Calibration Options Customizable settings are provided to help optimize the calibration process for the size and arrangement of specific capture volumes. Adjusting the values to those which best match the capture volume will help improve tracking performance.   Wanding Data Wanding data can be saved in CSV format for later use and reloaded using the Load and Save buttons.   Once the wand data has been collected, press the "Next" button to proceed.   4.2.4 Calculate Calibration Once a set of wand data has been captured in the previous step, it needs to be processed. The length of time taken for the calculation depends on the calibration quality level selected in the "Camera Preview" step. To start the calculation, press the "Start Calculation" button.   When the calculation is complete it will stop scrolling and display information about the level of quality obtained. The lower the numbers displayed for the Standard and Mean Error, the better. Typically it is desirable to have the Standard error below "0.5" and the Mean Error below "0.5". If the result of the calculation is unsatisfactory, consider returning to the "Wand Capture" stage to collect a new set of data.   Once a good calibration has been obtained, press the "Next" button to proceed.   4.2.5 Ground Plane The Ground Plane step allows you to identify the level of the physical floor (ground), set the scale of the coordinate system and how the axes of the 3D world are oriented. The cameras will be displayed in 3D using the positions calculated during the calibration step. The software will also reconstruct and display 3D coordinate data for markers detected in the scene. Until the ground plane is set, the position of the cameras and markers may not appear in the expected positions.   The Ground Plane is a 3-4-5 triangle with reflective markers in the three corners of the triangle. When the ground plane is placed within the capture volume, its three markers should show up in the 3D view with lines between them. When the triangle is detected and there are only 3 markers in the scene, press the "Set Ground Plane" button. After pressing the button, the cameras and markers will be updated to their new positions.     3D View   Rotate View : The 3D view can be rotated by pressing down the left mouse button and moving the mouse.   Translate View : The 3D view can be translated by pressing down the right mouse button and moving the mouse.   Translate View : The 3D view can be zoomed in or out by using the mouse-wheel or by pressing down the left and right mouse buttons and moving the mouse.   Ground Plane Calculation   Set Ground Plane : Press this button to update the calibration when the Ground Plane is in the desired position. It is only possible to press the button when 3 markers are in the scene, additional markers will cause the button to be disabled.   Ground Plane Size : Enter the size, in millimeters, of the "4" side of the Ground Plane 3-4-5 triangle. The "4" side is the Z axis. The default size is 300mm, 400mm, 500mm. It is important that this number is set accurately since it is used to scale the coordinate system for all tracking data. This number should be entered before pressing the "Set Ground Plane" button.   Height Offset : If the Ground Plane is not physically located on the exact floor (ground), then a height offset can be entered in millimeters to compensate for the difference. This number should be entered before pressing the "Set Ground Plane" button.   Point Cloud Tracking Options If needed, the parameters which control the reconstruction of 3D markers from 2D position data can be fine-tuned manually. Typically the default values should provide good results.   Once the Ground Plane has been set, proceed to the next step by pressing the "Next" button.   4.2.6 Results The calibration process is now complete. This last step provides a chance to review the results of the calibration and store them for later use. A number of tools are provided for visualizing the data and measuring the quality. If there is a custom application which will be making use of the Point Cloud COM object, it will need to load the calibration file saved at this stage in order to perform any tracking. It is also possible to load existing calibrations for inspection and use.         Camera Calibration Use the Save and Load buttons to store calibrations for later use or retrieve previous calibrations.   Capture Volume This tool provides a way to visualize the physical area of the capture volume to determine where there will be good tracking coverage. It can be adjusted to display the coverage area for all cameras, or to just show a subset where multiple cameras overlap.   Calibration Verification It is possible to measure the quality of the calibration by using a wand with two markers at a known distance apart (an accuracy wand). When the "Measure accuracy wand" box is checked, the wand can be moved through the capture volume and a real-time estimation of the error for the wands position will be displayed. The distance of the points on the wand should be entered in millimeters.   Camera Rays The Camera Rays tool can be used to determine which cameras are able to see which markers. It does this by drawing a line between every marker and the cameras which are detecting it.   Sample Data Capture 3D marker tracking data can be exported in CSV file format from this area by recording a capture sample and saving it. Use the button with the red circle to start recording and the button with the black square to stop. The "Save" button allows you to save the recorded data to a file. The exported data is formatted with the first column indicating the number of markers, the second column with a timestamp, and then 3 additional columns (x,y,z) per marker.   4.2.7 Options Several aspects of the Calibration tool interface can be customized to better suit your needs. The settings which control customization are found in the Options window, which can be launched by "Options" button in the lower-left corner of the Calibration tool. Changes to these settings will be saved when the Calibration tool is closed, and restored when it is started again.