These are for AR technology when it comes to marketing, this can be of use. But at the same time, we need to consider the fact that the user have to pull out their phone and point at an object.
That could actually limit the use case, since not everyone is willing to do that, and there is some concern about privacy. But what is the backbone of this technology?
How can we achieve markerless tracking?
This is a new step from marker-based tracking, which is powerful but still has a long way to go.
At the end of the day, we need some features to track, and in this case, it would be the image features itself. (and it seems like GPS can be also used to do augmented applications)
And the above software basically, does that extract the features from a given image store it for comparison in some other times.
The author of the code!
Basically rapid changes, the high frequency is changed in the image, and we want to have a lot of those since it is much better to track.
1. How to use subtle AR filters to survive your Zoom meetings?
2. The First No-Headset Virtual Monitor
But Neural networks can actually improve this matter as well.
Like the above image. (and images with low variance, or in other words changes are not a good suit for this application).
Not really sure what this is, but the author wanted to upgrade this branch to support WASM.
The main reason for this was to increase performance, and they have still yet to achieve the same level as other service platforms such as the 8th wall but things are coming along. (8th wall is a very good and powerful solution but pricey).
On mobile devices, we need some sort of video, this case would be the webcam. And stream the videos to the canvas.
There are some examples that I can study.
Wow, some crazy stuff is coming to life!