Skip to content

Parameter Settings

This chapter describes the components of the RVCManager parameter settings and how to adjust the parameters to obtain the desired image data. Frequently asked questions about parameter adjuatment can be found in Advanced Topics - FAQs about Capture Parameter Adjustment.

Judgement Basis for Capturing Effect

A point cloud with a good capture should meet the following conditions:

  • There are no obvious missing points in the point cloud or depth map, and the effective points have a high fill rate in the field of view.

  • There is no obvious flying noise on the point cloud.

  • The point cloud data has a high degree of reproduction of the captured object, and there is no obvious deformation or bending.

Capture Parameters

The Capture Parameters are mainly used to adjust the raw image and point cloud effects captured by the scanner, and the parameter options that can be set are related to the type of scanner connected.

Camera Mode

Monocular scanner can not choose camera mode, binocular scanner can choose Left Camera, Right Camera, Dual Camera Mode. The point cloud in Dual Camera Mode is the intersection of the point clouds captured by the left and right cameras, and its absolute accuracy is higher than that of the single camera mode, which is suitable for guidance applications with higher accuracy requirements; the capturing time in single camera mode is shorter. Different camera modes can be selected according to the capturing scene, accuracy and rhythm needs.

Left Camera Mode Right Camera Mode Dual Camera Mode
CameraModeLeft CameraModeRight CameraModeDouble

Capture Mode

Fast/Normal Mode: Available for RVC-P/I/X Series. The Fast Mode increases the scanner's capturing speed by reducing the number of projections, and its projection and processing time is 30% shorter than that of the Normal Mode.

Normal Mode Fast Mode
CaptureModeNormal CaptureModeFast

Ultra/Robust Mode: Available for RVC-G Series. The Robust Mode is based on the Ultra Mode to increase the number of images, and use a specific algorithm to filter out ambient light, which can effectively solve the problem of strong ambient light interference on the structured light, and improve the anti-interference ability.

AntiInter Reflection Mode: When there are multiple reflective surfaces in the scene, multiple reflections of the structured light will affect the imaging effect, resulting in a residual point cloud at the seams of the reflective surfaces. In this case, the previous capture modes can not meet the capturing needs. The AntiInter-Reflection Mode can be used to optimize the projection pattern, and reduce the structured light calculation errors caused by multiple reflections. AntiInter-Reflection Mode requires firmware upgrade (RVC-P/I/X Series: v2.6.5 or above, RVC-G Series: v1.1.0 or above) and does not support CUDA.

Normal Mode AntiInter Reflection Mode
AntiInterReflectionOff AntiInterReflectionOn

Projector Color

For RGB projector scanners, the Projector Color (Red/Green/Blue/White) can be switched in Dual Camera Mode to improve capturing results. This function is not available in single camera mode.

In general, an object reflects light with a wavelength similar to its own color and absorbs light that is complementary to its own color. Light source colors that are easily absorbed by objects are not recommended for capturing. Choosing the right color of light source can create enough contrast between the feature area and its surroundings to obtain more detailed features. The yellow workpiece shown below absorbs blue light, and the projection color is recommended to be set to red.

The object being captured is yellow Mono scanner + red projection Mono scanner + blue projection Mono scanner + white projection
ProjectorColorCapture ProjectorColorRed ProjectorColorBlue ProjectorColorWhite

2D Capture Settings

2D Capture parameters depend on the type of scanner. RVC Series Scanner come standard with mono cameras, please contact customer service if you need color cameras.

In general, when the 2D image is too dark, you can increase Exposure2D, Gain2D, and select Yes for ProjectorEnable2D; when the 2D image is too bright, you can decrease Exposure2D and Gain2D. Gamma2D is rarely used, so we recommend that you leave it at its default value.

Once the 2D parameters are set, take 2D capture or 3D capture to view the results of the 2D image. Take the calibration board below as an example, the capturing effect can be seen in the following table.

Caliboard
Normal capturing Too dark Too bright
Point cloud 2DRegularPointCloud 2DUnderexposurePointCloud 2DOverexposurePointCloud
2D image 2DRegularImage 2DUnderexposureImage 2DOverexposureImage

Detailed explanations and adjustment examples for each parameter are shown below.

2D Exposure time

Adjust the brightness of the 2D image by adjusting Exposure2D. The larger the parameter, the brighter the image. Under or overexposure can result in a loss of image detail. The range of 2D exposure times that can be set depends on the scanner model, and can be queried by the range displayed in the software interface.

2D exposure 30 ms 2D exposure 80 ms 2D exposure 100 ms
2DExposure30 2DExposure80 2DExposure100

2D Gain

Adjusts the brightness of a 2D image by controlling the sensor of the scanner. The higher the gain, the more sensitive it is to light.

2D gain 5 dB 2D gain 11 dB 2D gain 16 dB
2DGain5 2DGain11 2DGain16

2D Gamma

The Gamma correction expands the details of the dark tones of an image and corrects the deviation of the luminance; the Gamma value defaults to 1 and normally does not need to be adjusted. Normally, when the Gamma correction is less than 1, the highlights of the image are expanded and the dark tones are compressed; when the Gamma correction is greater than 1, the highlights of the image are compressed and the dark tones are expanded.

2D Gamma 0.5 2D Gamma 1 2D Gamma 1.5
2DGamma0p5 2DGamma1 2DGamma1p5

Projector Enable 2D

When Capturing 2D images, you can choose whether to turn on the light machine for fill-in projection. Capturing with ProjectorEnable2D on makes the 2D image brighter and has no effect on the point cloud.

ProjectorEnable2D off ProjectorEnable2D on
ProjectorOff ProjectorOn

3D Capture Settings

Normally, when point clouds are too dark, you can increase Exposure3D, Gain3D, Brightness, and select Yes for ProjectorEnable2D; when point clouds are too bright, you can decrease Exposure3D, Gain3D and Brightness. If there are large differences in the colors and materials of objects in the scene, or if there are reflective areas, you can capture in HDR mode.

Firstly, you can take a test capture with the default parameters. By looking at the solid color point cloud or depth map, you can quickly determine the missing point cloud. Blank areas in a solid color point cloud and black areas with NAN depth values in a depth map represent missing point clouds. Take the calibration board below as an example, the results can be seen in the table below.

Caliboard
Normal Capturing Underexposure Overexposure
Point cloud 3DRegularPointCloud 3DUnderexposurePointCloud 3DOverexposurePointCloud
Depth map 3DRegularDepthMap 3DUnderexposureDepthMap 3DOverexposureDepthMap

3D Exposure Time

Adjust the integrity of the point cloud and depth map by modifying the exposure time during 3D capturing. This needs to be adjusted according to the actual situation; underexposure or overexposure can result in missing point clouds.

The range of 3D exposure times that can be set depends on the scanner model. Typically, the 3D exposure time range is [3, 100] for P/I/X Series mono scanners, [11, 100] for color scanners, and [20, 100] for G Series laser scanners, which can be queried by the range displayed in the software interface.

3D exposure 11 ms 3D exposure 28 ms 3D exposure 100 ms
3DExposure11 3DExposure28 3DExposure100

3D Gain

The quality of the point cloud and depth map is adjusted by controlling the sensor of the scanner. Higher gain is more sensitive to light and also more sensitive to noise signals, which can increase the brightness of the point cloud, but may also result in an increase in point cloud noise.

3D gain 0 dB 3D gain 2 dB 3D gain 10 dB
3DGain0 3DGain2 3DGain10

Scan Times

This parameter, which applies only to RVC-G Series laser scanners, sets the number of scans when capturing in Robust Mode. The principle is to combine multiple image sequences obtained from multiple scans into a single image according to some rules and methods to improve image quality. In general, the higher the Scan Times, the greater the resistance to ambient light and the more image detail that can be obtained, but the capturing time will be extended.

Scan Times 2 Scan Times 4
ScanTime2 ScanTime4

Light Contrast

The Light Contrast is used to adjust the signal strength in the raw data captured by the scanner, and can be used to enhance the integrity of the point cloud and depth map when ambient or reflected light is strong. The default value is 3, which is not normally adjusted.

Light Contrast 0 Light Contrast 3 Light Contrast 10
Contrast0 Contrast3 Contrast10

Projector Brightness

Adjusts the Brightness value of the scanner's projection unit when taking pictures. The Brightness setting range is 1~240, and the default value is 240. the larger the value, the brighter the projection brightness.

Brightness 31 Brightness 83 Brightness 240
Brightness31 Brightness83 Brightness240

Band Width

Bandwidth is generally used in poor network conditions to reduce the impact of frame loss caused by unstable bandwidth. Setting range value 30%~100%, the default value is 100%. The smaller the bandwidth setting, the slower the transmission speed, which will result in longer capturing time.

HDR Parameters

The HDR function is mainly used to adjust the integrity of the point cloud and depth map, which can improve the quality of the point cloud of black light-absorbing and reflective objects. In scenarios where the material of the objects varies greatly in color, a single exposure may not be compatible with all objects, and the HDR function can provide additional 3D exposures to simultaneously capture objects of different colors and materials.The HDR parameter options are related to the scanner type.

The parameter HDR Extra Times has three options: 0, 1, and 2, which indicate no HDR, 1 additional 3D exposure, and 2 additional 3D exposures respectively. The 3D exposure time, 3D gain, scan times, and projector brightness of each exposure can be adjusted. For example, the following table shows the relationship between the exposure sequence and the corresponding HDR parameters of a laser scanner.

OperationHDR Extra TimesExposure sequenceExposure TimeGainScan TimesProjector Brightness
No HDR0Only 1 exposureSet directly in 3D Capture Settings
Use 1 additional 3D exposure1Initial exposureHDR Exposure [1]/msHDR Gain[1]HDR Scan Times[1]HDR Brightness[1]
First additional exposureHDR Exposure [2]/msHDR Gain[2]HDR Scan Times[2]HDR Brightness[2]
Use 2 additional 3D exposure2Initial exposureHDR Exposure [1]/msHDR Gain[1]HDR Scan Times[1]HDR Brightness[1]
First additional exposureHDR Exposure [2]/msHDR Gain[2]HDR Scan Times[2]HDR Brightness[2]
Second additional exposureHDR Exposure [3]/msHDR Gain[3]HDR Scan Times[3]HDR Brightness[3]

Take the following example of a scene containing a black light-absorbing object with a metallic reflective object.

HDRSample
Single exposure, black part underexposed Single exposure, overexposure of light colors and reflections 2 additional exposures using HDR function to synthesize a complete point cloud
HDRUnderexposure HDROverexposure HDRTwice

Post-Processing Parameters

The post-processing parameters are mainly used for post-processing the generated point cloud such as denoising, smoothing, downsampling, etc. for better viewing and application. The parameter options that can be set are related to the scanner type.

The denoising function includes confidence denoising, clustering denoising and edge denoising. The specific functions are as follows:

Confidence Denoising

By adjusting the Confidence Threshold, noise reduction can be performed according to the point cloud confidence level. This function is mainly for the cases of insufficient streak intensity and streak intensity error caused by multiple reflections. Users can first check the confidence range of the whole field of view in the display window, and then select the appropriate confidence threshold, below which the points will be filtered out. An example of the operation is shown below:

  1. In the point cloud of the metal part shown in the figure below, there is an abnormal flying-up noise at the round hole, which is inferred to be an error in the intensity of the stripes caused by reflections.

    ConfidenceDenoiseOff

  2. View the confidence map and find the approximate location of the reflection anomaly, the confidence level is 0.300. As a basis for judgment, you can set the confidence denoising threshold to more than 0.300, such as 0.310. Usually, this kind of noise corresponds to the darker color area in the confidence map.

    ConfidenceDenoiseMap

  3. Re-Capturing in 3D, the noise caused by reflections has been removed.

    ConfidenceDenoiseOn

Clustering Denoising

The clustering denoising function removes single noise or contiguous noise that fly up abnormally, with two parameters: Clustering Distance (clustering denoising distance threshold) and Clustering Points (clustering denoising effective points). The principle is: according to the density of the points to divide the original point cloud into a number of small areas (i.e., clustering), if the number of points in a small area is less, the area are judged as noise and removed.

  • Clustering Distance: Used to measure the density of points and cluster the point cloud. The smaller the Clustering Distance, the stronger the denoising effect, but if it is set too small, more valid points will be deleted by mistake. When the distance threshold is 0, no clustering denoising is performed.

  • Clustering Points: Used to determine whether the small area of the point cloud after clustering is a noise. The larger the Clustering Points, the stronger the denoising effect, but if it is set too large, more effective points will be deleted by mistake.

No clustering denoising Clustering denoising (Clustering Distance 0.258 mm, Clustering Points 100)
ClusterDenoiseOff ClusterDenoiseOn

Edge Denoising

The edge denoising function is only available in binocular mode, and its principle is to remove unreliable point clouds within the perimeter of invalid points, thus reducing the noise in the point cloud. The higher the Edge Denoising value, the stronger the denoising intensity, but it may cause the loss of detail information in the point cloud.

No edge denoising Edge Denoising level 2
EdgeDenoiseOff EdgeDenoiseOn

Smooth Level

Smooth Level is used to adjust the smoothness of the point cloud and is divided into 4 levels: Off, Weak, Medium and Strong, the default is Off. You can see the change of point cloud smoothing by observing the point cloud thickness.

Smoothing off Smooth Level Medium
SmoothnessOff SmoothnessMiddle

Downsample Distance

This parameter is used to downsample the point cloud. When Downsample Distance is smaller than the original point cloud spacing, the point cloud is basically unchanged. When Downsample Distance is larger than the original point cloud spacing, the point cloud spacing is approximately equal to the average downsampling spacing. The larger the Downsample Distance, the lesser the number of points.

No downsampling Downsample Distance 1 mm
Downsample-1 Downsample1

Calculate Normal Vector

The normal vector of the point cloud can be calculated by selecting "Yes" for Normal Calc Enable and setting Normal Calc Radius (the radius of the normal vector calculation).

DisplayNormalPointCloud

The normal vector of a point cloud can be approximated as a set of vertical lines on the object surface. The radius of normal vector calculation affects the accuracy of the normal vectors. If the radius is set too small, there are not enough points to calculate the local model and the result may be skewed; if the radius is set too large, the calculation process may be long and the result will be smooth.

Z-Truncate

Under the calibration board coordinate system or the camera coordinate system, the point cloud can be truncated according to a given Z-direction height range, which can quickly removing unwanted data outside the target area.

Raw point cloud Use Z-Truncate to quickly remove flying-up noise
ZTruncateOff ZTruncateOn

Bilateral Filtering

The bilateral filtering function is suitable for inspection applications that use edge features. If the edges of a point cloud become unsharp or even distorted after using the normal smoothing function, bilateral filtering can be used to improve the quality of the point cloud. The principle is to smooth the point cloud by adjusting the values of the bilateral filter convolution kernel size, depth and spatial Gaussian distribution, while still preserving features such as edges.

Bilateral filtering is only available in monocular mode and is not supported in the CUDA version. This function consists of 3 parameters: Bilateral Filter Kernal Size, Bilateral Filter Depth Sigma, and Bilateral Filter Space Sigma.

  • Bilateral Filter Kernal Size: Determines the range of pixels involved in the convolutional smoothing calculation. The larger the Kernal Size is, the more pixels are involved in the calculation, the more accurate the result is, but the speed will be slower.

  • Bilateral Filter Depth Sigma: Determines the depth range of the points to be added to the smoothing. The larger the Depth Sigma, the greater the depth difference between the points involved in smoothing. When capturing a thin workpiece on a plane, the depth distribution should not be set too large, otherwise the thin workpiece will be smoothed together with the points on the plane, resulting in blurring of the edge features.

  • Bilateral Filter Space Sigma: By adjusting the appropriate Space Sigma, points at different distances from the window center can be weighted. The larger the Space Sigma, the closer the smoothing result is to the mean value of all points in the window; the smaller the Space Sigma, the closer the smoothing result is to the value of the pixel point at the window center.

Raw point cloud Use normal smoothing function Use bilateral filtering
BilateralOff BilateralOffSmooth BilateralOn

Device Functions

The Device Functions option allows you to set parameters automatically according to the capturing requirements. Parameters that have already been set can be imported and exported, and can be called up directly for subsequent capturing. The parameter options that can be set depend on the type of scanner connected.

Set ROI

The Set ROI function is only available in single camera mode. When acquiring the point cloud of the object to be measured, the original field of view can be cropped to get a customized field of view, and the part not retained is no longer available. The selected field of view area is called the ROI (Region of Interest). By setting the ROI, the amount of data can be reduced and the data processing time can be lowered, thus improving the detection efficiency. The procedure is as follows:

  1. Click the button to set the ROI.

    ROIBefore_en

  2. Select the ROI region via the pop-up window, click [Confirm].

    ROI_en

  3. Capture again to get the point cloud, 2D image, depth map and confidence map of the cut area.

    ROIAfter

Auto White Balance

The Auto White Balance function determines the base color temperature of the image based on the selected Region of Interest, compared with 18% neutral gray, for color balance. This function is available only for color scanners. The Auto White Balance function can be used when the color of 2D images taken with a color camera is distorted.

  1. Click the button to perform Auto White Balance.

    WhiteBalanceBefore_en

  2. Select the ROI (Region of Interest) via the pop-up window.

    WhiteBalanceROI_en

    Click [Confirm] to get the 2D image after auto white balance.

    WhiteBalanceAfter

Auto HDR

The Auto HDR function automatically sets the HDR parameters according to the ROI selected. The HDR parameters will be automatically adjusted for the selected region, and the generated capture parameters will be written to the scanner for reuse. The HDR parameters that can be automatically set by this function include: HDR Extra Times, and the 3D exposure time, 3D gain, brightness for each exposure. The operation steps are as follows:

  1. Click the [Execute] button to execute Auto HDR.

    AutoHDRBefore_en

  2. Select the ROI (Region of Interest) via the pop-up window and click [Confirm].

    AutoHDRROI_en

  3. The HDR parameters will be automatically adjusted, and a 3D point cloud will be captured. You can check the parameter updates in the HDR Parameters.

    AutoHDRAfter_en

Auto Clustering Denoising

The clustering denoising function removes single or continuous noise from abnormal flying-ups, and also has an automatic adjustment option. Click the [Execute] button, the Clustreing Distance and Clustering Points will be automatically adjusted, and the adjusted parameters will be written to the scanner.

Abnormal flying-up noise Auto Clustering (Clustering Distance 0.237 mm, Clustering Points 40)
AutoClusterDenoiseOff AutoClusterDenoiseOn

Export Parameter File

In actual applications, sometimes it is necessary to set a fixed set of parameters for the project. After adjusting the parameters, you can export the capture parameters to a configuration file in .json format by using the Export Parameter File function.

Import Parameter File

When you need to call a saved parameter group, you can load the parameters from a configuration file to RVCManager through the Import Parameter File function.

Caution

Before importing parameters, you need to check whether the parameters in the configuration file are applicable to the currently connected scanner, otherwise an abnormality may occur. It is recommended to export/import parameters between scanners of the same model, and make sure the capturing environment is similar.

Common loading parameter exceptions are as follows:

  • Configuration file format error. Currently, only the .json format is supported.

  • Camera mode mismatch. Camera modes are selectable for binocular scanners, but not for monocular scanners. The configuration parameters of the two types of scanners are not common.

  • The P/I/X Series scanners are categorized into Normal/Fast Mode and the G Series laser scanners are categorized into Ultra/Robust mode. The configuration parameters of the two types are not common.

Calibration Parameters

Displays the camera calibration parameters, including External Parameter Matrix, Internal Parameter Matrix, and Distortion Parameter. If these parameters are not consistent with those shipped from the factory, it will affect the capturing results.

  • External Parameter Matrix: Parameters in the world coordinate system, such as the position of the camera, rotation direction, etc.. Divided into rotation matrix and translation matrix, the rotation matrix and translation matrix together describe how to transform the point from the world coordinate system to the camera coordinate system.

  • Internal Parameter Matrix: Parameters related to the camera's own characteristics, such as the camera's focal length and pixel size.

  • Distortion Parameter: The image distortion that occurs in image processing due to optical system deformation, reflective concave surfaces, and other factors during the capturing process. It is a numerical parameter that can be used to characterize image distortion, often using the inverse radius ratio Rd and the radius vector distribution coefficient K1.

CameraParameter_en

Detect Calibration Board

If there is a calibration marker in the capture scenario, it can be detected by this function. After completing the capturing, navigate to Detect CAL Board and click [Execute]. If the detection is successful, a pop-up window will appear to identify the marker position. The center of the detected marker is highlighted with a yellow-green cross in the 2D image, you can view and copy the 3D point position (3D Positions) and 2D pixel information (Pixel Coord) of the marker center through the data below.

CaliboardCenter_en

Coordinate Parameters

The coordinate parameters includes the calibration board coordinate system, customer coordinate system, and left/right camera coordinate system settings.

When you need to change the position and direction of the point cloud based on the coordinate axis, you can modify the base coordinate system of the data through the Coord Parameters. After the modification, the shape of the point cloud and the relative positions of the internal points will not be changed.

Calibration board coordinate system Camera coordinate system
DisplayCoordinateSystem CameraCoordinateSystem

Customer Coordinate System

Before use, make sure that the "Visible: Customer Coord" option is turned on in the Display Area. You can also turn on the "Visible: Coord" option to compare with the original coordinate system, to check if the coordinate system is transformed by the required amount.

  1. In "Coord Type", select the type of coordinate system to be transformed to (camera coordinate system/calibration board coordinate system/customer coordinate system, and left camera coordinate system/right camera coordinate system for a binocular scanner), and in "Customer Coord Base", select the original coordinate system before transformation.

  2. Input the X, Y, Z offset (DX, DY, DZ) and rotation (RZ, RY, RZ) of the coordinate system, or use the mouse to drag the slider to modify the values. For display purposes, the axes of the modified coordinate system are thicker and shorter than the original axes.

    UserCoordinateSystem1_en

  3. After the modification, the system will automatically calculate the converted coordinate system according to the conversion matrix. When you execute 3D capturing, the Display Area will show the position and direction of the point cloud under the customer coordinate system.

    UserCoordinateSystem2_en

Customer Coordinate System Quick Configuration

In some applications, it is necessary to use the surface of the captured sample as the reference plane to construct the customer coordinate system. When it is difficult to adjust the parameters manually, you can select 3 points in the point cloud to determine the origin, the positive direction of the X-axis, and the plane of the Y-axis one by one, so as to quickly determine the customer coordinate system. Set the parameters of the coordinate system to quickly fit the plane of the coordinate system.

  1. Click the [Execute] button, the Display Area prompts for the selected point.

    QuickSetCoordinateSystem1_en

  2. Select the origin, X-axis direction and Y-axis direction in the point cloud, and the coordinates of each point are displayed in the lower left corner.

    QuickSetCoordinateSystem2_en

  3. After selecting points, the Display Area shows the adjusted customer coordinate system if the setting is effective, and the parameter setting area updates the customer coordinate system parameters.

    QuickSetCoordinateSystem3_en

  4. When a 3D capture is taken, the position and orientation of the point cloud is transformed to the customer coordinate system.

    QuickSetCoordinateSystem4_en