You simply need to use a DPI sufficient to accurately represent the image that was processed by the TMS - so that the same features recognized by the TMS are apparent in the image.

What do you mean by ..

Quote:

Because I noticed that on some points that the photos online feramenta for generating marker identifies almost any point in the picture which hinders the subsequent recognition.

Is this when you use a low DPI and the image appears jagged?

I meant in relation to the assessment tool to generate the DAT file (tracker) is the image, as some others they identify as weak as strong. What criterion is used to make the image evaluation as a good marker?

Normally a good contrast between two pixel. I think the TMS works similar!

Robert

The TMS is?

andercid wrote:

The TMS is?

That's the shortcut for the trackable creation process.

My Trackables -> New Project -> ...

To recognize a person I'm having difficulty, is something that should be done to facilitate the identification of person (marker)? I want to display the data of the person when she is recognized but the tool evaluates the marker as weak. there is anything I can do?

You can try increasing the contrast of the picture using image processing software. That may help, but an image of a person's face may not have enough sharp corners to make a good target. You could also try adding extra feature points around the face. For example, if the photo has the person posing against a solid background, try adding a background with a random pattern to introduce more features. Use the feedback from the TMS to evaluate your progress. The stars are one metric, but also look at the distribution of features points across your image to see where the holes are.

- Kim