ESP-VISION Programming Guide

[中文]

ESP-VISION is a Low-Code Edge AI & Computer Vision Framework for Espressif SoCs. It deeply integrates essential capabilities including camera capture, image processing, video encoding and decoding, network transmission, model deployment, and AI inference, while providing unified and standardized Python APIs that enable developers to rapidly build edge applications combining visual capture, intelligent recognition, display, and media streaming.

This guide is organized as follows.

• Introduction
  • What Is ESP-VISION
  • Quick Experience
  • Key Features
  • Supported Boards
• Chip and Board Support
  • Supported Development Boards
  • Chip Versus Board Capabilities
  • Standard MicroPython Features
  • WebREPL and EV-MUX Network Control
• Get Started
  • Prerequisites
  • Build, Flash, and Monitor
  • Common idf.py Commands
  • Application-only Flashing
  • Run Your First Script
• Concepts
  • MicroPython Language Fundamentals
  • Packages and Deployment
  • The Image Model
  • Image Processing
  • AI Inference
  • The Camera Pipeline
  • Codecs and Streaming
  • Startup Sequence
• Storage, Filesystems and Data Protection
  • Media and filesystem ownership
  • Optional volumes and automatic mounting
  • USB with two volumes
  • Failure policy: preserve data and keep control available
  • One writer, safe publication and power loss
  • Backup and restore policy
  • Explicit recovery of an unmounted Flash volume
  • Firmware installation and acceptance
• API Reference
  • MicroPython APIs
  • sensor – Camera
  • image – Image Processing
  • display – LCD Output
  • lcd_touch – Touch Input
  • espdl – Model Inference
  • tflite – Model Inference
  • image.ImageIO – Image Stream
  • uvc — USB webcam output
  • storage – Diagnostics and Confirmed Recovery
• How-To Guides
  • Build vision applications with AI
  • Add a New Python Module
  • Contribution and Review
  • Customize Firmware Features
  • Publish qualified firmware
  • Introduce a New Model
  • Train an ESPDet Pico Model
  • Add a New Board
  • EV-MUX v4 Host Integration Entry
  • Agent Code Generation Guide
• Solution Architecture
  • Layered Overview
  • Capture-to-Output Data Flow
  • Communication Subsystem
  • Source Tree
  • Board Composition
  • Chip-Dependent Sources
  • MicroPython Overlay
• Communication System and EV-MUX
  • The complete path at a glance
  • Recommended reading order
  • Six distinct concepts
  • Current implementation boundaries
• Device Core host SDK
  • Application API reference
  • Advanced export API reference
  • Integrating the Device Core npm Package
  • Device management CLI
  • Installation and compatibility
  • Architecture and starting point
  • Licensing
• Project Relationships
  • ESP-VISION, MicroPython, and OpenMV
  • Dependency Layers
• Licensing
  • License Inventory
  • How to Interpret the Licenses
  • OpenMV and GPL Code Paths
  • Adding Third-Party Code