Remove Noise from Surface

Description

This Step is used to remove noise from the surface data based on the similarity of neighboring data points.

Workflow

The process of configuring this Step is shown below:

Configure the input. Connect the Step ports in the graphical programming workspace or select the input under Input in the parameter configuration panel.
Determine whether to use a feature region.
Set other parameters.
Select the desired output items under Output. For an expandable output item, click ▶ and configure the Min and Max values to determine the acceptable range for the item.
Run the Step and view output.

Parameters

Parameter	Description
Use Feature Region	Unselected: The Step processes all of the surface data. Selected: The Step processes the surface data defined by the feature region(s). In this case, set the following parameters: Feature Region Mode Use this parameter to decide whether to process the data inside or outside of the feature region(s). Feature Region See Feature Region for details on setting and adjusting a feature region.
Use Intensity Image	When unselected, the Step processes the depth map; when selected, the Step processes the intensity image.
Depth Difference Threshold	Determines whether two neighboring points are similar. If the depth value difference between two neighboring points is less than or equal to the threshold, they are considered similar and grouped into the same class.
Intensity Difference Threshold	Only visible when Use Intensity Image is selected. Determines whether two neighboring points are similar. If the intensity value difference between two neighboring points is less than or equal to the threshold, they are considered similar and grouped into the same class.
Neighborhood	Determines which points around a point are considered its neighboring points. 4-neighbor: The points directly above, below, to the left, and to the right of a point are its four neighbors. 8-neighbor: A point’s horizontal and vertical neighbors (as in 4-neighbor) and its four diagonal neighbors are its eight neighbors.
Min Number of Points in a Class	After clustering, if the number of data points in a class is lower than the set minimum, all points in that class will be considered noise and removed.
Use Statistical Filter	When noise points are connected to the main data and the differences in height or intensity values are small, use the statistical filter to further remove noise. After selecting this parameter, you need to set Standard Deviation Calculation Method and Standard Deviation Multiplier.
Standard Deviation Calculation Method	Based on actual values: Calculates the standard deviation based on actual depth or intensity values. Based on relative values: Calculates the standard deviation based on relative depth or intensity values.
Number of Neighbors	Only visible when the “Standard Deviation Calculation Method” is set to Based on relative values. Determines how many neighbors of a point are used to calculate the differences in depth or intensity values. The mean of these differences is assigned to the point, thus updating the depth or intensity values of all data points in a class as relative values.
Standard Deviation Multiplier	Determines how many standard deviations a data point is allowed to deviate from the mean. Points outside the allowable deviation range will be removed as noise. A smaller values means more points will be considered noise and removed.

Parameter

Description

Use Feature Region

Unselected: The Step processes all of the surface data.

Selected: The Step processes the surface data defined by the feature region(s). In this case, set the following parameters:

Feature Region Mode

Use this parameter to decide whether to process the data inside or outside of the feature region(s).
Feature Region

See Feature Region for details on setting and adjusting a feature region.

Use Intensity Image

When unselected, the Step processes the depth map; when selected, the Step processes the intensity image.

Depth Difference Threshold

Determines whether two neighboring points are similar. If the depth value difference between two neighboring points is less than or equal to the threshold, they are considered similar and grouped into the same class.

Intensity Difference Threshold

Only visible when Use Intensity Image is selected.

Determines whether two neighboring points are similar. If the intensity value difference between two neighboring points is less than or equal to the threshold, they are considered similar and grouped into the same class.

Neighborhood

Determines which points around a point are considered its neighboring points.

4-neighbor: The points directly above, below, to the left, and to the right of a point are its four neighbors.
8-neighbor: A point’s horizontal and vertical neighbors (as in 4-neighbor) and its four diagonal neighbors are its eight neighbors.

Min Number of Points in a Class

After clustering, if the number of data points in a class is lower than the set minimum, all points in that class will be considered noise and removed.

Use Statistical Filter

When noise points are connected to the main data and the differences in height or intensity values are small, use the statistical filter to further remove noise.

After selecting this parameter, you need to set Standard Deviation Calculation Method and Standard Deviation Multiplier.

Standard Deviation Calculation Method

Based on actual values: Calculates the standard deviation based on actual depth or intensity values.
Based on relative values: Calculates the standard deviation based on relative depth or intensity values.

Number of Neighbors

Only visible when the “Standard Deviation Calculation Method” is set to Based on relative values.

Determines how many neighbors of a point are used to calculate the differences in depth or intensity values. The mean of these differences is assigned to the point, thus updating the depth or intensity values of all data points in a class as relative values.

Standard Deviation Multiplier

Determines how many standard deviations a data point is allowed to deviate from the mean. Points outside the allowable deviation range will be removed as noise. A smaller values means more points will be considered noise and removed.

Output Description

The output of this Step is processed surface data that can be used as input into other Steps.

Troubleshooting

For common errors, see Error Code List.
If the error code is inconsistent with the error message, please contact Technical Support for help.

CV-W4001

Error: The “Depth Difference Threshold” value must be greater than 0. Please enter a valid value.

Solution: Make sure the parameter value is greater than 0.

CV-W4002

Error: The set “Neighborhood” is invalid. Please select a valid option from the drop-down list.

Solution: Select a valid neighborhood type from the drop-down list.

CV-W4003

Error: The “Min Number of Points in a Class” value must exceed 0. Please enter a valid value.

Solution: Make sure the parameter value is greater than 0.

CV-W4004

Error: The set “Standard Deviation Calculation Method” is invalid. Please select a valid option from the drop-down list.

Solution: Select a valid standard deviation calculation method from the drop-down list.

CV-W4005

Error: The “Number of Neighbors” value must be greater than 0 and less than the “Min Number of Points in a Class.” Please enter a valid value.

Solution: Ensure the parameter value is greater than 0 and less than the “Min Number of Points in a Class.”

CV-W4006

Error: The “Standard Deviation Multiplier” value must be greater than 0. Please enter a valid value.

Solution: Ensure the parameter value is greater than 0.

Is this page helpful?

Leave a feedback

Thanks for your support!

You can give a feedback in any of the following ways:

We Value Your Privacy

We use cookies to provide you with the best possible experience on our website. By continuing to use the site, you acknowledge that you agree to the use of cookies. If you decline, a single cookie will be used to ensure you're not tracked or remembered when you visit this website.

I accept

I don't accept