ImFusion::BlobDetector Class Reference

#include <ImFusion/Vision/BlobDetector.h>

Detects locally connected regions ("blobs") in grayscale or RGB images. More...

Detailed Description

Detects locally connected regions ("blobs") in grayscale or RGB images.

The detector sweeps a range of intensity thresholds on a grayscale version of the input image. At each threshold it binarizes the image, extracts connected components, and applies optional geometric and color filters. Components that appear at similar locations across multiple consecutive thresholds are merged; only blobs observed at least minRepeatability times are returned. Stable detections are reported as sub-pixel centers with an estimated diameter and the contour from the strongest observation.

This implementation follows the multi-threshold blob detection approach used in OpenCV's SimpleBlobDetector, including inertia- and convexity-based filtering.

Classes
struct	Params
	Configuration parameters for BlobDetector. More...
struct	Blob
	Result of blob detection for a single stable region. More...

Public Member Functions
	BlobDetector (const BlobDetector::Params &parameters=BlobDetector::Params())
	Construct a blob detector with the given parameters.
void	detect (const TypedImage< unsigned char > *image, std::vector< Blob > &keypoints) const
	Detect blobs in image and append results to keypoints.

Protected Member Functions
void	findBlobs (TypedImage< unsigned char > *image, TypedImage< unsigned char > *binaryImage, std::vector< Blob > &centers) const
	Extract blob candidates from a single binarized image at the current threshold.

Member Function Documentation

detect()

void ImFusion::BlobDetector::detect	(	const TypedImage< unsigned char > *	image,
		std::vector< Blob > &	keypoints ) const

Detect blobs in image and append results to keypoints.

image may be single-channel grayscale or three-channel RGB; RGB input is converted to grayscale internally. On success, keypoints is cleared and filled with stable blobs that satisfy the configured filters and repeatability criteria.

Parameters

image	Input image. Must not be null.
keypoints	Output vector receiving detected blobs.

---

The documentation for this class was generated from the following file:

• ImFusion/Vision/BlobDetector.h