Overlap Filter

Class: OverlapBlockV1

Source: inference.core.workflows.core_steps.analytics.overlap.v1.OverlapBlockV1

The OverlapFilter is an analytics block that filters out objects overlapping instances of some other class

For instance, for filtering people on bicycles, "bicycle" could be used as the overlap class.

Examples applications: people in a car, items on a pallet

The filter will remove the overlap class from the results, and only return the objects that overlap it. So in the case above, bicycle will also be removed from the results.

Type identifier

Use the following identifier in step "type" field: roboflow_core/overlap@v1to add the block as as step in your workflow.

Properties

Name	Type	Description	Refs
name	str	Enter a unique identifier for this step..	❌
overlap_type	str	Select center for centerpoint overlap, any for any overlap.	❌
overlap_class_name	str	Overlap Class Name.	❌

The Refs column marks possibility to parametrise the property with dynamic values available in workflow runtime. See Bindings for more info.

Available Connections

Compatible Blocks

Check what blocks you can connect to Overlap Filter in version v1.

• inputs: Path Deviation, Dynamic Crop, Object Detection Model, Moondream2, Byte Tracker, Segment Anything 2 Model, Perspective Correction, Detections Stabilizer, Detections Consensus, Detections Stitch, Line Counter, Time in Zone, Time in Zone, YOLO-World Model, Detections Merge, Overlap Filter, Instance Segmentation Model, Google Vision OCR, PTZ Tracking (ONVIF).md), Detections Transformation, Object Detection Model, Detection Offset, Detections Filter, Byte Tracker, VLM as Detector, Velocity, Instance Segmentation Model, Dynamic Zone, Byte Tracker, Detections Classes Replacement, Path Deviation, VLM as Detector, Bounding Rectangle, Template Matching
• outputs: Path Deviation, Dynamic Crop, Label Visualization, Color Visualization, Pixelate Visualization, Corner Visualization, Roboflow Custom Metadata, Distance Measurement, Byte Tracker, Segment Anything 2 Model, Florence-2 Model, Circle Visualization, Roboflow Dataset Upload, Model Comparison Visualization, Perspective Correction, Detections Stabilizer, Detections Stitch, Detections Consensus, Blur Visualization, Line Counter, Time in Zone, Trace Visualization, Time in Zone, Florence-2 Model, Bounding Box Visualization, Crop Visualization, Stitch OCR Detections, Detections Merge, Overlap Filter, PTZ Tracking (ONVIF).md), Size Measurement, Detections Transformation, Line Counter, Detection Offset, Detections Filter, Roboflow Dataset Upload, Ellipse Visualization, Triangle Visualization, Byte Tracker, Model Monitoring Inference Aggregator, Dot Visualization, Velocity, Byte Tracker, Path Deviation, Detections Classes Replacement, Background Color Visualization

Input and Output Bindings

The available connections depend on its binding kinds. Check what binding kinds Overlap Filter in version v1 has.

Bindings

• input

  • predictions (Union[instance_segmentation_prediction, object_detection_prediction]): Object predictions.

• output

  • overlaps (object_detection_prediction): Prediction with detected bounding boxes in form of sv.Detections(...) object.

Example JSON definition of step Overlap Filter in version v1

{
    "name": "<your_step_name_here>",
    "type": "roboflow_core/overlap@v1",
    "predictions": "$steps.object_detection_model.predictions",
    "overlap_type": "Center Overlap",
    "overlap_class_name": "<block_does_not_provide_example>"
}