We have high quality goods images (from 1,000 to 10,000 images per project, examples attached to project).
On other hand, we have store shelves pictures, made by phone camera.
We need to develop software, which will be able to find all entires from provided data set on photo and give all their positions. Pay attention, that objects can be rotated, have multiple entries and so on. Check example picture in attachment.
- you can use any language to develop;
- goods database don't change at all;
- one photo can be processed up to 60 sec;
- you can use any interface to get inout images, for example you can get path to local stored image from command line execution or http link to image as GET param;
Please describe your experience with opencv and preferred language.
Also, please, provide estimate for first prototype to test and for final product, even if it will be dirty estimate.