![]() |
ImFusion C++ SDK 4.5.0
|
#include <ImFusion/Vision/BlobDetector.h>
Configuration parameters for BlobDetector. More...
Configuration parameters for BlobDetector.
Public Attributes | |
| float | thresholdStep |
| Step size in intensity units between successive threshold values in the range [minThreshold, maxThreshold). | |
| float | minThreshold |
| Lowest grayscale intensity used for binarization during the multi-threshold sweep. | |
| float | maxThreshold |
| Upper bound (exclusive) of the threshold sweep. | |
| size_t | minRepeatability |
| Minimum number of threshold levels at which a blob must be redetected at roughly the same location before it is accepted in the final output. | |
| float | minDistBetweenBlobs |
| Minimum spatial separation in pixels between blob centers when grouping detections across thresholds. | |
| bool | filterByColor |
| If true, reject blobs whose binarized pixel value at the center differs from blobColor. | |
| unsigned char | blobColor |
| Expected binarized pixel value (0 = dark blob on bright background, 255 = bright blob on dark background) used when filterByColor is enabled. | |
| bool | filterByArea |
| If true, reject blobs whose contour area in pixels is outside [minArea, maxArea). | |
| float | minArea |
| Minimum contour area in pixels². | |
| float | maxArea |
| Maximum contour area in pixels² (exclusive). | |
| bool | filterByCircularity |
If true, reject blobs whose circularity 4π·area/perimeter² falls outside [minCircularity, maxCircularity). | |
| float | minCircularity |
| Minimum allowed circularity in [0, 1]. Values near 1 indicate round blobs. | |
| float | maxCircularity |
| Maximum allowed circularity (exclusive). | |
| bool | filterByInertia |
| If true, reject blobs whose inertia ratio (ratio of minimum to maximum second central moment eigenvalues) falls outside [minInertiaRatio, maxInertiaRatio). | |
| float | minInertiaRatio |
| Minimum allowed inertia ratio in [0, 1]. | |
| float | maxInertiaRatio |
| Maximum allowed inertia ratio (exclusive). | |
| bool | filterByConvexity |
| If true, reject blobs whose convexity (contour area divided by convex-hull area) falls outside [minConvexity, maxConvexity). | |
| float | minConvexity |
| Minimum allowed convexity in (0, 1]. Values near 1 indicate compact, convex shapes. | |
| float | maxConvexity |
| Maximum allowed convexity (exclusive). | |
| float ImFusion::BlobDetector::Params::thresholdStep |
Step size in intensity units between successive threshold values in the range [minThreshold, maxThreshold).
Smaller steps increase sensitivity to weak blobs but also runtime and the number of intermediate detections.
| float ImFusion::BlobDetector::Params::maxThreshold |
Upper bound (exclusive) of the threshold sweep.
Thresholds satisfy minThreshold <= t < maxThreshold.
| size_t ImFusion::BlobDetector::Params::minRepeatability |
Minimum number of threshold levels at which a blob must be redetected at roughly the same location before it is accepted in the final output.
Higher values suppress unstable detections that appear only at a single threshold.
| float ImFusion::BlobDetector::Params::minDistBetweenBlobs |
Minimum spatial separation in pixels between blob centers when grouping detections across thresholds.
Also compared against blob radius so that overlapping blobs are merged into the same track rather than reported twice.
| float ImFusion::BlobDetector::Params::minArea |
Minimum contour area in pixels².
Blobs smaller than this are discarded when filterByArea is enabled.
| float ImFusion::BlobDetector::Params::maxArea |
Maximum contour area in pixels² (exclusive).
Blobs with area greater than or equal to this value are discarded when filterByArea is enabled.
| bool ImFusion::BlobDetector::Params::filterByInertia |
If true, reject blobs whose inertia ratio (ratio of minimum to maximum second central moment eigenvalues) falls outside [minInertiaRatio, maxInertiaRatio).
Low ratios indicate elongated shapes.
| float ImFusion::BlobDetector::Params::minInertiaRatio |
Minimum allowed inertia ratio in [0, 1].
Also scales the confidence weight used when averaging sub-threshold observations of the same blob.