Skip to content

3D Square Probe

Operator Function

Search for edge points in a specific direction, returning geometric parameters such as points, lines, planes, regions, etc.

Parameter Introduction

Input Parameters

ParameterRangeDefault ValueDescriptionIllustration
Input Image0-80The IM number for image input

Calculation Parameters

ParameterRangeDefault ValueDescriptionIllustration
Region2D Window/BoxBoxManually select an appropriate ROI region
Can bind to select existing ROI regions
Direction+X/-X/+Y/-Y/+Z/ -Z/Select All PointsSelect All PointsSelect search direction
+X: Search along positive X-axis
-X: Search along negative X-axis
+Y: Search along positive Y-axis
-Y: Search along negative Y-axis
+Z: Search along positive Z-axis
-Z: Search along negative Z-axis
Select All Points: Select all points within ROI
Exclusion Cone Angle45.000The cone angle when searching along the selected direction; points outside the cone will be preserved
Resolution1000The reciprocal of step length when searching along the selected direction; increasing resolution means more sampling points, but also increases operator runtime.
Search Foreground OnlyWhether to only search for foreground edge points; if checked, manually set Z-direction threshold to separate foreground and background.
Output GeometryPoint/Line/Plane/SegmentLinePoint: Centroid coordinate point of point cloud region
Line: Contains center point of line segment and direction vector of line
Plane: Contains geometric center point of plane and normal vector of plane
Segment: Contains coordinates of two endpoints of line segment
Filter MethodNo Filter/Statistical Filter/RANSAC FilterNo FilterSee Filter Methods

Result Display

ParameterRangeDefault ValueDescriptionIllustration
Point Settrue/falsefalseTarget point set; if enabled, displayed in the image
Pointtrue/falsefalsePoints found by the tool; if enabled, displayed in the image
Linetrue/falsefalseLines found by the tool; if enabled, displayed in the image
Planetrue/falsefalsePlanes found by the tool; if enabled, displayed in the image
ROI Point Settrue/falsefalsePoint set of ROI selected region; if enabled, displayed in the image
Segmenttrue/falsefalseLine segments found by the tool; if enabled, displayed in the image
Segment Center Pointtrue/falsefalseMidpoints of line segments found by the tool; if enabled, displayed in the image

Output Parameters

NameTypeRangeDescription
Resultbooltrue/falsetrue for success
false for failure
Warningbooltrue/falsetrue indicates there is a warning
false indicates there is none
MessagestringOutputs success, error, or warning information; if there is no error or warning, it is empty
TimefloatOperator execution time, unit:ms
regionpointsetPoint set region
middle_pointpoint3dMidpoint coordinates
Lineline3dLine information
planeplanePlane information
Segmentsegment3dLine segment information
Segment Centerpoint3dLine segment midpoint information

Filter Methods

ParameterRangeDefault ValueDescriptionIllustration
No FilterNo filtering of found edge points
Statistical FilterThis filtering method first performs a best-fit on the found edge points, then uses the standard deviation multiplication factor to calculate the distance threshold. Calculate the distance from all edge points to the fitted model; if this distance exceeds the threshold, it is considered an outlier.
Standard Deviation Multiplier: Recommended range [1.0, 2.0]
RANSACThis filtering method consists of two steps: hypothesis and verification. The hypothesis step randomly selects n points from the point set for geometric model fitting; the verification step uses the remaining points to calculate their distance to this geometric model. Points with distances less than the threshold are called inliers. If the number of inliers meets the threshold, the geometric model from this hypothesis step is considered the best model, and iteration stops.
Distance Threshold: If the distance from a point to the output geometry exceeds this threshold, it will be filtered out.
Iteration Count: If iteration exceeds this count, the algorithm stops.

Exception Troubleshooting

No.Exception InformationCorresponding ParameterSolution
1The input value is {0}, invalid region typeRegion TypeOnly supports window2d, box, pointset
2Region is empty1. Check if input point cloud is empty
2. Check if ROI region encloses any point cloud
3. Check if bound pointset is empty
3The input value is {0}, invalid search directionSearch DirectionOnly supports +X, -X, +Y, -Y, +Z, -Z, no probing
4Failed to search edge points1. Check if search direction is correct
2. Check if resolution is negative; if so, change to positive
5The input value is {0}, invalid filter methodFilter MethodOnly supports No dispersion filter, Statistical outlier removal, RANSAC
6The input value is {0}, invalid output geometry typeGeometry TypeOnly supports Point, Line, Plane, Segment 3D

Example Introduction

Engineering Design

  1. Select the Load 3D Point Cloud tool to load the 3D point cloud image to be processed into IM0.

  2. Select the 3D Point Cloud Crop tool, copy IM0 to IM1, IM2, IM3 respectively for different operations.

  3. Select the 3D Square Probe tool.

Tool Usage

  1. Select the input image for operation; the image number must match the IM number where the image is located in the project.

  2. Select the region type as Box, move the box to the position to be measured, enclosing the point cloud to be measured.

Usage Tips

  1. Use the ROI controller on the image window to drag or scale the box;
  2. Directly modify the start or end coordinates of the box in the calculation parameters to adjust the box position and size;
  1. Set parameters

Select multiple 3D Square Probe tools, set output geometry to Point and Line respectively with search directions +X, +Y, -X, -Y, and set output geometry to Plane with different regions selected.

  1. Check the content you want to display in the result display section

  2. Click Test to check if the image window and parameters meet expectations

  3. If there are no issues, click Save, then run the operator in the run list to view the results in the corresponding IM

Local Image

AI-Vision, Making 3D Measurement Easier