I currently have a web service that does as follows:
– User inputs image
– Web service launches a C program using OpenCV
– The program loads training data
– The program detects and recognises certain objects
– The program records these detections and recognitions in a MySQL database
– If the program has detected and recognised objects, it adds it to the training data
This program can take up to 1 minute to run per image. How could I structure Amazon Web Services to do this efficiently?
Coincidentally, AWS released a detailed tutorial about processing images, using EC2, S3, SimpleDB, and SQS. It should get you most of the way there.