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