Detections Stitch

Class: DetectionsStitchBlockV1

Source: inference.core.workflows.core_steps.fusion.detections_stitch.v1.DetectionsStitchBlockV1

This block merges detections that were inferred for multiple sub-parts of the same input image into single detection.

Block may be helpful in the following scenarios: * to apply Slicing Adaptive Inference (SAHI) technique, as a final step of procedure, which involves Image Slicer block and model block at previous stages. * to merge together detections performed by precise, high-resolution model applied as secondary model after coarse detection is performed in the first stage and Dynamic Crop is applied later.

Type identifier

Use the following identifier in step "type" field: roboflow_core/detections_stitch@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_filtering_strategy	str	Which strategy to employ when filtering overlapping boxes. None does nothing, NMS discards lower-confidence detections, NMM combines them..	✅
iou_threshold	float	Minimum overlap threshold between boxes. If intersection over union (IoU) is above this ratio, discard or merge the lower confidence box..	✅

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 Detections Stitch in version v1.

• inputs: Blur Visualization, Triangle Visualization, Anthropic Claude, Trace Visualization, YOLO-World Model, Label Visualization, LMM, Model Monitoring Inference Aggregator, Roboflow Dataset Upload, Time in Zone, Absolute Static Crop, Image Preprocessing, Relative Static Crop, Byte Tracker, Image Threshold, Reference Path Visualization, Segment Anything 2 Model, Slack Notification, Stability AI Outpainting, Instance Segmentation Model, SIFT, Roboflow Dataset Upload, Identify Outliers, Detection Offset, Google Vision OCR, Stability AI Inpainting, Background Color Visualization, Byte Tracker, Detections Stabilizer, Detections Transformation, CSV Formatter, Circle Visualization, Image Blur, Keypoint Visualization, VLM as Detector, Google Gemini, OpenAI, Image Convert Grayscale, PTZ Tracking (ONVIF).md), Line Counter Visualization, Model Comparison Visualization, Path Deviation, Dynamic Zone, Roboflow Custom Metadata, Image Slicer, Line Counter, Crop Visualization, Corner Visualization, Stitch OCR Detections, Multi-Label Classification Model, Pixelate Visualization, Local File Sink, Image Slicer, Velocity, Overlap Filter, Mask Visualization, VLM as Classifier, Color Visualization, Clip Comparison, Polygon Visualization, Email Notification, Keypoint Detection Model, Perspective Correction, Detections Consensus, Path Deviation, Moondream2, Bounding Rectangle, Camera Calibration, OpenAI, Detections Filter, Detections Merge, Bounding Box Visualization, Camera Focus, Detections Stitch, Instance Segmentation Model, CogVLM, OpenAI, Twilio SMS Notification, Dynamic Crop, Detections Classes Replacement, Depth Estimation, Halo Visualization, Florence-2 Model, Dot Visualization, Template Matching, Classification Label Visualization, Time in Zone, Webhook Sink, SIFT Comparison, Stability AI Image Generation, VLM as Detector, Florence-2 Model, LMM For Classification, Object Detection Model, Ellipse Visualization, Image Contours, Llama 3.2 Vision, Single-Label Classification Model, Identify Changes, Grid Visualization, Stitch Images, OCR Model, Byte Tracker, Object Detection Model, Polygon Zone Visualization
• outputs: Blur Visualization, Triangle Visualization, Polygon Visualization, Trace Visualization, Size Measurement, Label Visualization, Distance Measurement, Perspective Correction, Model Monitoring Inference Aggregator, Path Deviation, Detections Consensus, Roboflow Dataset Upload, Bounding Rectangle, Time in Zone, Detections Filter, Detections Merge, Byte Tracker, Bounding Box Visualization, Segment Anything 2 Model, Roboflow Dataset Upload, Detections Stitch, Detection Offset, Detections Classes Replacement, Dynamic Crop, Halo Visualization, Byte Tracker, Background Color Visualization, Stability AI Inpainting, Detections Stabilizer, Detections Transformation, Florence-2 Model, Dot Visualization, Time in Zone, Circle Visualization, Florence-2 Model, PTZ Tracking (ONVIF).md), Path Deviation, Model Comparison Visualization, Dynamic Zone, Ellipse Visualization, Roboflow Custom Metadata, Line Counter, Stitch OCR Detections, Crop Visualization, Corner Visualization, Line Counter, Pixelate Visualization, Velocity, Byte Tracker, Mask Visualization, Overlap Filter, Color Visualization

Input and Output Bindings

The available connections depend on its binding kinds. Check what binding kinds Detections Stitch in version v1 has.

Bindings

• input

  • reference_image (image): Original image that was cropped to produce the predictions..
  • predictions (Union[object_detection_prediction, instance_segmentation_prediction]): Model predictions to be merged into the original image..
  • overlap_filtering_strategy (string): Which strategy to employ when filtering overlapping boxes. None does nothing, NMS discards lower-confidence detections, NMM combines them..
  • iou_threshold (float_zero_to_one): Minimum overlap threshold between boxes. If intersection over union (IoU) is above this ratio, discard or merge the lower confidence box..

• output

  • predictions (Union[object_detection_prediction, instance_segmentation_prediction]): Prediction with detected bounding boxes in form of sv.Detections(...) object if object_detection_prediction or Prediction with detected bounding boxes and segmentation masks in form of sv.Detections(...) object if instance_segmentation_prediction.

Example JSON definition of step Detections Stitch in version v1

{
    "name": "<your_step_name_here>",
    "type": "roboflow_core/detections_stitch@v1",
    "reference_image": "$inputs.image",
    "predictions": "$steps.my_object_detection_model.predictions",
    "overlap_filtering_strategy": "nms",
    "iou_threshold": 0.4
}