Back to Adaptive Vision Library website

# FitPathToEdges_Direct

Performs a series of 1D edge detections and creates a path from the detected points.

Applications: Tracing of an object contour, whose rough location and shape is known beforehand.

### Syntax

C++
C#

void avl::FitPathToEdges_Direct
(
const avl::Image& inImage,
const avl::PathFittingField& inFittingField,
atl::Optional<const avl::CoordinateSystem2D&> inFittingFieldAlignment,
atl::Optional<float> inScanStep,
int inScanWidth,
const avl::SamplingParams& inSamplingParams,
const avl::EdgeScanParams& inEdgeScanParams,
avl::Selection::Type inEdgeSelection,
atl::Optional<const avl::LocalBlindness&> inLocalBlindness,
atl::Optional<int> inMaxInterpolationLength,
atl::Optional<float> inMaxDeviationDelta,
float inMaxIncompleteness,
atl::Conditional<avl::Path>& outPath,
atl::Optional<atl::Array<atl::Conditional<avl::Edge1D> >&> outEdges = atl::NIL,
atl::Optional<atl::Conditional<avl::Profile>&> outDeviationProfile = atl::NIL,
atl::Optional<avl::PathFittingField&> outAlignedFittingField = atl::NIL,
atl::Array<avl::Segment2D>& diagScanSegments,
atl::Array<avl::Rectangle2D>& diagSamplingAreas,
atl::Array<avl::Profile>& diagBrightnessProfiles,
atl::Array<avl::Profile>& diagResponseProfiles
)


### Parameters

Name Type Range Default Description
inImage const Image& Image to fit the path to
inFittingField const PathFittingField& Path fitting field
inFittingFieldAlignment Optional<const CoordinateSystem2D&> NIL Adjusts the fitting field to the position of the inspected object
inScanStep Optional<float> 0.0 - 5.0f Optional implicit conversion of the input path to an equidistant one
inScanWidth int 1 - 5 The width of each scan field (in pixels)
inSamplingParams const SamplingParams& Parameters controlling the sampling process
inEdgeScanParams const EdgeScanParams& EdgeScanParams ( ProfileInterpolation: Quadratic4 SmoothingStdDev: 0.6f MinMagnitude: 5.0f EdgeTransition: BrightToDark ) Parameters controlling the edge extraction process
inEdgeSelection Selection::Type Selection mode of edges
inLocalBlindness Optional<const LocalBlindness&> NIL Defines conditions in which weaker edges can be detected in the vicinity of stronger edges
inMaxInterpolationLength Optional<int> 0 - 1 Maximal number of consecutive points not found
inMaxDeviationDelta Optional<float> 0.0 - NIL Maximal difference between deviations of consecutive path points
inMaxIncompleteness float 0.0 - 0.999 0.1f Maximal fraction of edge points not found
outPath Conditional<Path>& Fitted path or nothing if the fitting failed
outEdges Optional<Array<Conditional<Edge1D> >&> NIL Found edges
outDeviationProfile Optional<Conditional<Profile>&> NIL Profile of distances between the actual path points and the corresponding reference path points
outAlignedFittingField Optional<PathFittingField&> NIL Fitting field used; in the image coordinate system
diagScanSegments Array<Segment2D>& Segments along which the scans were run
diagSamplingAreas Array<Rectangle2D>& Areas from which the input image is sampled
diagBrightnessProfiles Array<Profile>& Extracted image profiles
diagResponseProfiles Array<Profile>& Profiles of the edge (derivative) operator response

### Optional Outputs

The computation of following outputs can be switched off by passing value atl::NIL to these parameters: outEdges, outDeviationProfile, outAlignedFittingField.

### Description

The operation tries to fit a given path to edges present in the inImage image. Internally, it performs a series of scans with the ScanSingleEdge_Direct filter along a number of specific scan segments which length is always equal to the inFittingField width and cannot be less than 4. The found points are then used to determine the actual position of the path in the image. Only inMaxIncompleteness fraction of these scans may fail. If the fitting according to the given parameters is not possible, outPath is set to Nil.

There are also another parameters that control the path fitting process. The inMaxDeviationDelta parameter defines the maximal allowed difference between deviations of consecutive points from the input path points. If some of the scans fail or if some of found points are classified to be wrong according to another control parameters, output path points corresponding to them are interpolated depending on points in their nearest vicinity. No more than inMaxInterpolationLength consecutive points can be interpolated. The exception to this behavior are points which were not found on both ends of the input path. Those are not part of the result at all.

### Hints

• Connect an input image to the inImage input.
• Define inEdgeScanParams.EdgeTransition to detect a particular edge type, and only that type.
• If no or too few edge points are found, try decreasing inEdgeScanParams.MinMagnitude.
• If some of the scans may fail, set the inMaxIncompleteness input accordingly.
• Use the outEdges outputs to visualize the scanning results.

### Examples

Fitting a path to the edges of a plastic capsule
(inEdgeSelection = First, inEdgeScanParams.EdgeTransition = BrightToDark).

### Remarks

Read more about Local Coordinate Systems in Machine Vision Guide: Local Coordinate Systems.

This filter is a part of the Shape Fitting toolset. To read more about this technique, one can refer to the Shape Fitting chapter of our Machine Vision Guide

### Hardware Acceleration

This operation supports automatic parallelization for multicore and multiprocessor systems.