How to Choose a Good Target Image for Tracking in AR - Part 1

As an AR developer, I have worked on countless number of image tracking AR applications. Common questions are: Is this image okay? Why is the tracking so bad? I'm going to answer all these questions once and for all!

This article series explains the fundamental principles in AR Image Tracking and hopefully helps you choose a target image wisely in your next project. It's recommended for anyone working in Augmented Images.

Which AR engine?

First of all, the concepts introduced in the series apply to all AR engines. It doesn't matter whether you are using Pictarize, 8th Wall, SparkAR, Vuforia or any other tools that provide AR image tracking. Each of these engines have their own implementations and optimizations for sure, but fundamentally they are all doing the same thing.

Overview of the series

The articles are designed to be read in order for a coherent and complete understandings. However, each individual parts are understandable by itself, so it's okay to skip if you see fit.

Part 2 - Fundamental Theory of Image Tracking

Part 2 is about theory. It explains how image tracking work deep down from image pixels level. However, don't be scared. it's not a super technical computer vision review. It's written in a way that ordinary people can understand.

Part 3 - Practical Examples of Good and Bad Target Images

Part 3 is a practical chapter. It discusses high level characteristics of images that are important to AR tracking. At the end, you will know how to choose a good target image, or modify one to make it good.

Part 4 - Target Image Analyzer

In part 4, we will introduce a tool that can automatically analyze the trackability of images. It's free and extremely easy to use. Basically, you just need to upload an image, and the program will return quantitative scores.

Next

Part 2 - Fundamental Theory of Image Tracking