# Optimizing Image Target Tracking

To ensure the highest quality image target tracking experience, be sure to follow these guidelines when selecting an image target.

### Preparation <a href="#preparation" id="preparation"></a>

Before training the *Image Targets*, you must prepare the images in digital format. The available formats are

* .jpg/jpeg
* .png
* .bmp

A good *Image Target* will have certain visual features that enables accurate pose estimation and stable tracking :

<table data-header-hidden><thead><tr><th width="345"></th><th></th></tr></thead><tbody><tr><td>Properties</td><td>Explanation</td></tr><tr><td><strong>Distinctive pattern</strong></td><td>A texture with evenly distributed and distinctive visual features (e.g. corners, blobs)</td></tr><tr><td><strong>No repetitive pattern</strong></td><td>Textures with no similar patterns repeating in space</td></tr><tr><td><strong>High contrast</strong></td><td>Image with high contrast</td></tr><tr><td><strong>Non-reflective</strong></td><td>Print on a matte surface that will not reflect light when printed</td></tr></tbody></table>

### Target Image Best Practices: <a href="#h_278610ccac" id="h_278610ccac"></a>

Image Targets that possess the following attributes will enable the best detection and tracking performance from the the SQUARS viewer. A basic rule is that complex and detailed images on well lit flat surfaces work best.

Examples of good target images:

<figure><img src="https://files.gitbook.com/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FRy95QTuRyUx2bNx9Zcjo%2Fuploads%2FuIMxwD0eHdBZI1iMJdOj%2FExamples%20of%20good%20target%20images%20(1).png?alt=media&#x26;token=52dd874e-7c0a-4001-8a15-b58940c1f53e" alt=""><figcaption></figcaption></figure>

Examples of bad target images:

<figure><img src="https://files.gitbook.com/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FRy95QTuRyUx2bNx9Zcjo%2Fuploads%2FQP4ZRsu5NADORbTJrMgo%2FExamples%20of%20bad%20target%20images%20(1).png?alt=media&#x26;token=cf3aa9e4-2e45-4864-ba4a-ff402ab0e0df" alt=""><figcaption></figcaption></figure>

### Target Rating <a href="#target-star-rating" id="target-star-rating"></a>

Image Targets are detected by analyzing the inherent features extracted from the target image and comparing them in real-time with the features in the live camera image. Targets are given a rating of either 'good', 'normal', or 'bad'. Although targets with **lower ratings ('bad' or 'normal') can still be detected and tracked**, it is advisable to use a 'good' rating target.

<figure><img src="https://files.gitbook.com/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FRy95QTuRyUx2bNx9Zcjo%2Fuploads%2FHbd27gjcgDtwQtFqAdhm%2FTarget%20Rating.png?alt=media&#x26;token=286a2963-9827-44c1-ad6d-6d00fae0ff9e" alt=""><figcaption></figcaption></figure>

To create a target that is accurately detected, you should use images according to the above attributes for an ideal Image Target.