A quick demo project exploring the text detection feature in Vision.framework
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
Vision Text Detection Demo.xcodeproj
Vision Text Detection Demo
.gitignore
LICENSE
README.md
demo.gif

README.md

Vision.framework Text Detection Demo

intro

Introduction

Vision is one of the many exciting new frameworks introduced at WWDC 17. Enhanced Text detection is among the many new features in Vision.framework. But since Apple hasn't provided any demo code on the Vision text detection, here is a quick sample project.

This demo app:

  1. takes a live video feed, detects text regions (VNTextObservation) and draws a green box around each region
  2. draws blue boxes around each character box detected (VNRectangleObservation)

Requirements

Running this app requires at least Xcode 9 beta 1 and a physical device running iOS 11 beta 1.

Other Notes

There is no OCR framework in the iOS SDK still. As a result, a seperate OCR engine is needed to recognize detected text content.