Skill Registration and Usage

🎯 What This Tool Does

visual understanding platform. This tool enables you to:

• - Register Skills - Register detection skills from platform locally
• Invoke Skills - Call registered skills with images or video frames
• Visualize Results - Draw bounding boxes, generate grid references, preview ROI/Tripwire
• Define Detection Areas - Use interactive workflows to define ROI (electronic fencing) or Tripwire (detection lines)

Supported Detection Types: Person detection, pedestrian counting, vehicle recognition, OCR, pose estimation, object tracking, etc.

🔧 Prerequisites

Get API Key

1. 1. Activate trial package

📚 Usage Guide

Basic Workflow

CODEBLOCK0

Or register a skill first then use it:

CODEBLOCK1

Define Detection Areas

Need to define regions of interest (ROI) or detection lines (Tripwire)?

• - ROI Workflow — Create electronic fencing, detect only in specified regions
• Tripwire Workflow — Draw detection lines, count crossing events

Both workflows include complete interactive steps and example dialogs.

View Complete Documentation

• - Type Definitions — Detection, ROI, Tripwire data structures
• Visualization Guide — Display detection results, generate grid
• Video Frame Extraction — Extract frames from video for detection
• Grid Input System — Use grid coordinates to specify points
• Example Code — Complete workflow examples

💡 Common Tasks

Register a Skill

After obtaining an ep-id:

Invoke a Skill

CODEBLOCK2

With ROI:
CODEBLOCK3

With Tripwire:
CODEBLOCK4

Preview ROI/Tripwire

Preview on image before invoking:

CODEBLOCK5

Generate Grid

Help users specify point locations using grid coordinates:

CODEBLOCK6