The A5000 ViewerCognex VisionPro 9.22 SR1

The 3D-A5000 sensor supports a set of custom properties that control acquisition parameters such as an exposure time, a region of interest, and more.

For many new 3D applications, Cognex recommends you use the A5000 Viewer to determine and set which acquisition parameters work best for your 3D-A5000 sensor. The Viewer supports a graphical user interface that can make it easier to configure the acquisition properties of a 3D-A5000 sensor, and VisionPro can pass the settings you choose to QuickBuild and/or your Visual Studio application automatically.

See the topic Custom Properties for details on modifying the custom properties using QuickBuild or programatically.

See the following sections for more information:

Launch the Viewer and Connect to a 3D Sensor

Launch the A5000 Viewer through the Start menu or double-click its icon on the Windows desktop:

Acquisition_LionSensor_Theory_PointClouds_ViewerIcon

The Viewer offers a single panel for configuring many acquisition settings and viewing their impact on the cloud points you acquire:

Acquisition_LionSensor_Theory_PointClouds_ViewerWithArrows

Be aware a 3D-A5000 sensor must have a configured IP address relative to the GigE Vision network adapter installed in your computer. Without the IP address VisionPro (including the A5000 Viewer) cannot communicate with the sensor. See the section GigE Vision Configuration Tool for more information.

Use the File menu to connect the application to a connected 3D-A5000 sensor.

Acquisition_LionSensor_Theory_PointClouds_FileMenu

Connect (or Alt+C) Connect the A5000 Viewer to the listed 3D-A5000 sensor.
Disconnect (or Alt+D) The Viewer prompts you to save any unsaved settings when you disconnect.
Export (or Alt+E) For each image acquisition, the Viewer offers you the ability to save each point cloud and/or the 2D intensity images from the separate GigE Vision cameras. Configure the following settings for saving the images:
  • Directory: Path of the root directory where to store the images
  • Index: An integer value used as a postfix in the name for each image
  • Point Clouds checkbox: Save the point cloud to the location specified by Directory using the value of Index in the name of the file. By default the Viewer uses the string "cloud-" in the name of the file but you can specify a different string. Save the list of 3D point coordinates in one of three file formats:
    • ASC: ASCII plain text
    • PCD: Point Cloud Data, used inside the Point Cloud Library (PCL)
    • PLY: Polygon File Format (a.k.a. Standford Triangle Format
  • Save: Save the point cloud and/or 2D intensity images of the last acquisition
  • Start Autosave: Automatically export each acquired point cloud and/or 2D intensity images and increment the value of Index after each successful export.
  • Intensity images checkbox: Save the greyscale intensity images to the location specified by Directory using the value of Index in the name of the file.
Acquisition Parameters

Acquisition_LionSensor_Theory_PointClouds_AcqParameters

Image countSpecify the of images used for 3D-reconstruction for each image acquisition
  • Higher values require longer measurement and processing times, but generate 3D range images with smaller variations of brightness.
  • Minimum value of 12 and maximum value of 60
ExposureChoose an exposure time for each 2D image, in microseconds
  • Choose a value so that the feaures under inspection appear as bright as possible in the 2D images without being overexposed. Examine the acquired 2D images on the Cameras tab of the viewport to see if they are overexposed as shown in the following example:
    Acquisition_LionSensor_Theory_PointClouds_OverExposed


    Click Auto for a baseline exposure setting optimized for the whole 3D scene. Manually fine-tune the value for the area of interest in your images.
  • Higher Exposure times can lead to slower acquisition times. The value can range from 10 to 5000 microseconds.
High Dynamic Range (HDR) settings Enable the High Dynamic Range (HDR) exposure settings in applications where your production environment can cause dark shadows and bright glare over the object being inspected. Enabling one of these makes the sensor acquire an additional sequence with another exposure time to use for reconstruction.
  • Enabling HDR low exposure captures a set of images with a shorter exposure time.

    You might use a low exposure setting if your object contains highly reflective components.
  • Enabling HDR high exposure captures a set of images with a higher exposure time.

    You might use a high exposure setting if your object contains diffuse components.

Be aware of the following when using HDR exposure settings:

  • Using HDR exposure settings will double or triple the acquisition time for each acquisition.
  • Be aware HDR exposure settings must be disabled to use continuous acquisition mode, such as the VisionPro Free Run trigger model.
CompressionEnable Low Compression to perform 10-bit acquisition and capture a larger dynamic range of pixel values. Be aware that Compression requires approximately a 4x longer exposure to take advantage of the larger dynamic range.

Use Compression as an alternative to using low and/or high HDR exposure settings. Compression has less impact on acquisition time but the results might not be as precise.

Region of Interest

You can specify a smaller camera resolution to limit the size of the image as required by your vision application. For example, if the object under inspection is significantly smaller than the default field of view for the sensor, acquisition and processing speeds will generally increase as you specify smaller parameters for the width, height and offset of your 3D range image.

Acquisition_LionSensor_Theory_PointClouds_NewWidthHeight

Use the following parameters to set the region of interest:

Acquisition_LionSensor_Theory_PointClouds_ROIParameters

WidthImage resolution along the x-axis

Valid values between 80 and 1440, by increments of 8
HeightImage resolution along the y-axis

Valid values between 4 and 1080, by increments of 4
2 x 2 BinningCombines 2 x 2 matrices of pixels into one pixel, generating significantly brighter images compared to those without 2 x 2 Binning enabled. 2 x 2 Binning offers the following benefits:
  • Provides higher sensitivity in images with dark features, allowing you to use a lower Exposure setting while achieving the desired brightness in the images your application captures
  • Produces a quarter of the points for a point cloud acquired without 2 x 2 Binning, which results in faster processing times but with half the lateral image resolution
  • Generates images with less image noise, offering better 3D data quality
2 x 2 Binning offers the maximum frame rate for your 3D-A5000 sensor based on your Exposure setting. A setting or 1300 microseconds or less results in the maximum frame rate of 500 Hz. The frame rate decreases with larger values for Exposure, down to the default of 276 Hz in full resolution (2 x 2 Binning is not enabled)
Offset x and Offset yUse Offset values to move the region of interest for either GigE Vision camera:
  • Offset x: Valid values between 0 and (1440 - width)
  • Offset y: Valid values between 0 and (1080 - height)
Reconstruction Quality

Choose a Reconstruction Quality control to choose a value between 0 (lowest quality, lowest reconstruction time) and 9 (highest quality, highest reconstruction time).

Acquisition_LionSensor_Theory_PointClouds_Reconstruction

Experiment with this parameter for best results.

Filtering

Choose a Pre-Filtering value applied to the GigE Vision cameras before reconstruction:

Acquisition_LionSensor_Theory_PointClouds_Prefiltering
  • Choose a value between 0 (minimum filter intensity) and 9 (maximum filter intensity).
  • The higher the setting, the lower the reconstruction time.
  • Pre-filtering removes noisy data such areas of the image that are overexposed, underexposed, or show low-pattern visibility

Choose Post-Filtering values to control the amount of data in the 3D reconstruction:

Acquisition_LionSensor_Theory_PointClouds_Postfiltering
Score thresholdEnter a minimum quality score each 3D point has to exceed before it will be included in the output 3D range image.

Increasing this value will lead to fewer outliers being included in the final 3D range image but will increase the number of missing points.

Cognex recommends a value of at least 0.7 and typically below 0.95 for most applications.
Outlier filteringThe options are Disabled, Permissive, Balanced and Strict, in that order. As you increase the policy from one to the next, the number of outliers included in the 3D range image decrease while increasing the risk of filtering out valid 3D points.
Working volumeDefine the volume of space within the field of view of both GigE Vision cameras in which 3D points will be constructed:
  • Standard: The limited volume with the highest accuracy for the 3D sensor
  • Extended: A greater volume of space but at the cost of some accuracy in the 3D points

Click Show Working Volume to toggle the view of the working volume in the 3D display.

Volume of Interest

By default the 3D sensor generates a 3D range image from all data within the field of view of both cameras. Enable a Volume of Interest and use the values X limits, Y limits and Z limits to extract a portion of the available 3D object volume. The settings allow you configure the smallest and highest limits for each axis.

Acquisition_LionSensor_Theory_PointClouds_VolumeOfInterest

Click From Point Cloud to limit the volume of interest for future image acquisitions based on the content of the current 3D reconstruction.

Click Visualize to enable a blue wire-frame box to represent the volume of interest in the 3D Viewer window.

Triggers

Configure the application to use a software trigger through the Viewer interface or a hardware trigger in response to a signal over the hardware trigger line.

Acquisition_LionSensor_Theory_PointClouds_Triggers

Be aware you must click Start before the application responds to any trigger events. Click Stop to prevent the application from responding to trigger events.

Hardware triggerThe application responds to a signal received over the hardware trigger line. For more information on hardware triggers, refer to the 3D-A5000 Series Sensors Hardware Reference, installed with your VisionPro installation or available from the VisionPro Support site.
Software triggerThe application performs an image acquisition each time you press Trigger.

Enable the Repeat checkbox repeat software trigger events at an interval you specify, in milliseconds (ms).
Continuous triggerThe application acquires images continuously without responding to trigger events (software or hardware). Continuous mode provides the highest possible rate of 3D image acquisitions.

The Viewer also displays a Trigger timeout to represent the minimum duration the 3D-A5000 requires between triggers (software or hardware).

3D Viewport

View each image acquisition in the 3D viewport:

Acquisition_LionSensor_Theory_PointClouds_Viewport

Left-click and drag the mouse inside the viewport to rotate the image. Click on a single face of the axis cube at the bottom-right to snap the view to that axis.

See the following sections for more information on using the Viewport:

Image Selection

The tabs along the top of the 3D viewport allow you to view the 2D intensity images used to create the point cloud, the point cloud itself, or the 3D range image calculated from the point cloud. The following figure shows the 2D intensity images from the latest acquisition:

Acquisition_LionSensor_Theory_PointClouds_CameraRow

Color

Use the Color pulldown menu to choose how to view the point cloud:

  • Greyscale: View the point cloud with an overlay of the grey values from the GigE Vision cameras.
  • Height: View the point cloud in color representing the height of features in the image.

    Click Configure to determine if the viewport assigns colors automatically or displays colors based on a range of Z values:

    Acquisition_LionSensor_Theory_PointClouds_ConfigureColor
    Click From Point Cloud to let the viewport assign a range of colors based on the volume of the point cloud along the z-axis.

  • None: View the point cloud with no greyscale overlay.
Geometry

Choose between Mesh or Points for how the viewport displays the point cloud.

Configurations

Viewer saves all configuration parameters in the file config.yml in the directory %APPDATA%\Cognex Corporation\VisionPro-A5000\. Use the Sensor menu to modify the current configuration file:

Acquisition_LionSensor_Theory_PointClouds_SensorMenu

Load ConfigurationReplace the current sensor configuration with the stored configuration from the configuration file.
Save ConfigurationStore the current sensor configuration to the configuration file.
Clear ConfigurationDeletes the stored configuration entries of the currently connected sensor from the configuration file.