Computationally Intensive Task on Google Glass

Question

What is the best way of going about running a computationally intensive task on Google Glass. The specific task I have in mind involves OpenCV and a lot of image processing. I single image is provided as input, and OpenCV analyzes the image. A desktop takes a couple seconds to complete this task, while Glass takes upwards of a couple minutes. What would be the best way to defer this processing to either a phone or computer? My code is in Java.

The way I see it there are two possibilities:

1. Defer the computationally intensive task to a phone or computer by sending the input image via Bluetooth.

2. Defer the computationally intensive task to a phone or computer by sending the input image through the cloud.

What would be the best solution for this particular problem? If I were to use the cloud approach, what cloud services would be the best to use will Google Glass?

Answer 1

Push the data via Mirror API upload methods as described here in order not to deal with transfer problems. Write an app server to run your algorithms and run it on Google App Server or your own server. Return results to Mirror API and make it return it as callback value to Google Glass.