I am very new to AR, ARCore and ARKit. The problem I am trying to solve is measure distance between visual features that are determined by AI algorithms. I want to use AR for measuring.
I am approaching this problem like this:
- AR captures the scene with the features points
- 2d image is extracted for identifying required visual features with Neural Network
- map 2d point to 3d coordinate in the same scene to create anchor
- calculate the distance between the anchors
I am very confused about the following things:
- Can we save an AR session and get the image that can be used as input for Neural Networks?
- If yes for 1, how to transfer the 2d coordinates obtained after a prediction from the neural network in ARCore or ARKit session for creating the anchor as the depth information is not there?
What is the best approach to solving this problem? If there are any other possible solutions to this problem, help is appreciated.
