Depth Estimation¶
Class: DepthEstimationBlockV1
Source: inference.core.workflows.core_steps.models.foundation.depth_estimation.v1.DepthEstimationBlockV1
🎯 This workflow block performs depth estimation on images using Apple's DepthPro model. It analyzes the spatial relationships
and depth information in images to create a depth map where:
📊 Each pixel's value represents its relative distance from the camera
🔍 Lower values (darker colors) indicate closer objects
🔭 Higher values (lighter colors) indicate further objects
The model outputs:
1. 🗺️ A depth map showing the relative distances of objects in the scene
2. 📐 The camera's field of view (in degrees)
3. 🔬 The camera's focal length
This is particularly useful for:
- 🏗️ Understanding 3D structure from 2D images
- 🎨 Creating depth-aware visualizations
- 📏 Analyzing spatial relationships in scenes
- 🕶️ Applications in augmented reality and 3D reconstruction
⚡ The model runs efficiently on Apple Silicon (M1-M4) using Metal Performance Shaders (MPS) for accelerated inference.
Type identifier¶
Use the following identifier in step "type" field: roboflow_core/depth_estimation@v1to add the block as
as step in your workflow.
Properties¶
| Name | Type | Description | Refs |
|---|---|---|---|
name |
str |
Enter a unique identifier for this step.. | ❌ |
model_version |
str |
The Depth Estimation model to be used for inference.. | ✅ |
The Refs column marks possibility to parametrise the property with dynamic values available
in workflow runtime. See Bindings for more info.
Runtime compatibility¶
-
hard— runtimeself_hosted_cpu; executionlocal - Requires a GPU; run_locally() loads a model that needs CUDA.
Available Connections¶
Compatible Blocks
Check what blocks you can connect to Depth Estimation in version v1.
- inputs:
Image Slicer,Polygon Zone Visualization,VLM As Classifier,Contrast Enhancement,Google Gemma API,MoonshotAI Kimi,Stability AI Image Generation,Image Threshold,Line Counter Visualization,Trace Visualization,Stitch OCR Detections,Camera Calibration,QR Code Generator,Anthropic Claude,Icon Visualization,SIFT Comparison,Morphological Transformation,S3 Sink,Color Visualization,LMM For Classification,Perspective Correction,Microsoft SQL Server Sink,Corner Visualization,Roboflow Custom Metadata,Google Vision OCR,Twilio SMS Notification,Halo Visualization,Image Blur,Morphological Transformation,Qwen-VL,Camera Focus,Email Notification,Roboflow Vision Events,Halo Visualization,Stability AI Inpainting,Classification Label Visualization,Stitch OCR Detections,Google Gemma,Event Writer,Grid Visualization,Qwen3.5-VL,Background Color Visualization,Mask Visualization,Llama 3.2 Vision,Ellipse Visualization,Email Notification,Reference Path Visualization,Image Slicer,Label Visualization,Twilio SMS/MMS Notification,Text Display,OPC UA Writer Sink,Dot Visualization,Polygon Visualization,Crop Visualization,Dynamic Crop,Absolute Static Crop,Circle Visualization,Image Preprocessing,Llama 3.2 Vision,Model Monitoring Inference Aggregator,Relative Static Crop,Camera Focus,OpenRouter,OpenAI,Florence-2 Model,OpenAI-Compatible LLM,MoonshotAI Kimi,Heatmap Visualization,Single-Label Classification Model,OpenAI,OCR Model,CogVLM,Blur Visualization,Depth Estimation,Instance Segmentation Model,Stability AI Outpainting,Anthropic Claude,Google Gemini,Qwen 3.6 API,Clip Comparison,Google Gemini,Background Subtraction,Keypoint Visualization,CSV Formatter,Webhook Sink,Bounding Box Visualization,Multi-Label Classification Model,LMM,OpenAI,Stitch Images,Florence-2 Model,Image Convert Grayscale,Current Time,Contrast Equalization,OpenAI,VLM As Detector,Google Gemini,Roboflow Visual Search,Triangle Visualization,Slack Notification,EasyOCR,Roboflow Dataset Upload,Pixelate Visualization,Roboflow Dataset Upload,PLC Writer,SIFT,Qwen 3.5 API,Anthropic Claude,Object Detection Model,Local File Sink,MQTT Writer,Image Contours,Polygon Visualization,Keypoint Detection Model,GLM-OCR,Model Comparison Visualization,Roboflow Asset Library Attributes - outputs:
VLM As Classifier,MoonshotAI Kimi,Stability AI Image Generation,Trace Visualization,Qwen2.5-VL,Image Stack,Anthropic Claude,Icon Visualization,SIFT Comparison,Morphological Transformation,Color Visualization,SmolVLM2,LMM For Classification,Single-Label Classification Model,Perspective Correction,Corner Visualization,Clip Comparison,Halo Visualization,Qwen-VL,Keypoint Detection Model,Halo Visualization,Object Detection Model,Google Gemma,Background Color Visualization,Ellipse Visualization,Email Notification,Twilio SMS/MMS Notification,Text Display,Polygon Visualization,Crop Visualization,Absolute Static Crop,Image Preprocessing,Template Matching,Relative Static Crop,OpenRouter,OpenAI,Florence-2 Model,VLM As Detector,OpenAI,Motion Detection,Heatmap Visualization,OCR Model,Perception Encoder Embedding Model,Blur Visualization,Barcode Detection,Depth Estimation,Instance Segmentation Model,Stability AI Outpainting,Anthropic Claude,YOLO-World Model,Google Gemini,Clip Comparison,Google Gemini,Background Subtraction,Keypoint Visualization,Buffer,Stitch Images,Florence-2 Model,Contrast Equalization,Mask Edge Snap,OpenAI,Qwen3-VL,Moondream2,VLM As Detector,Google Gemini,Triangle Visualization,CLIP Embedding Model,Detections Stabilizer,SIFT,Multi-Label Classification Model,Image Contours,Keypoint Detection Model,VLM As Classifier,Pixel Color Count,GLM-OCR,Image Slicer,Polygon Zone Visualization,Contrast Enhancement,Google Gemma API,Time in Zone,Semantic Segmentation Model,Image Threshold,Line Counter Visualization,Semantic Segmentation Model,Multi-Label Classification Model,Camera Calibration,ByteTrack Tracker,Google Vision OCR,Image Blur,Morphological Transformation,Camera Focus,Roboflow Vision Events,Stability AI Inpainting,Classification Label Visualization,SAM2 Video Tracker,Event Writer,Qwen3.5-VL,Mask Visualization,Llama 3.2 Vision,Dominant Color,Reference Path Visualization,Image Slicer,Label Visualization,Byte Tracker,SIFT Comparison,Dot Visualization,Dynamic Crop,Detections Stitch,Circle Visualization,Llama 3.2 Vision,BoT-SORT Tracker,SAM3 Video Tracker,Camera Focus,Gaze Detection,Segment Anything 2 Model,MoonshotAI Kimi,Single-Label Classification Model,QR Code Detection,Qwen3.5,CogVLM,Object Detection Model,SAM 3 Interactive,Qwen 3.6 API,Bounding Box Visualization,Multi-Label Classification Model,LMM,OpenAI,SAM 3,Image Convert Grayscale,Instance Segmentation Model,EasyOCR,Roboflow Visual Search,Roboflow Dataset Upload,SAM 3,Instance Segmentation Model,Keypoint Detection Model,Pixelate Visualization,Roboflow Dataset Upload,SORT Tracker,Instance Segmentation Model,Track Class Lock,Qwen 3.5 API,Object Detection Model,Anthropic Claude,Polygon Visualization,OC-SORT Tracker,SAM 3,Model Comparison Visualization,Single-Label Classification Model,Seg Preview
Input and Output Bindings¶
The available connections depend on its binding kinds. Check what binding kinds
Depth Estimation in version v1 has.
Bindings
-
input
-
output
image(image): Image in workflows.normalized_depth(numpy_array): Numpy array.
Example JSON definition of step Depth Estimation in version v1
{
"name": "<your_step_name_here>",
"type": "roboflow_core/depth_estimation@v1",
"images": "$inputs.image",
"model_version": "depth-anything-v2/small"
}