Image-based software designed to let brands know the exposure they are getting on social media can also be used to collate and curate images.

Indian-developed gazeMetrix launched last year following funding on startup website AngelList. It uses proprietary computer vision and machine learning algorithms to analyse user-created photos initially on Instagram, and has plans to approach Pinterest, Facebook and Twitter. It claims to analyse more than 12 million public Instagram photos a day.

The idea is that brand owners can see latest photos, with the most popular selected automatically, potentially for sharing on a fan page, Pinterest board or Twitter stream. An important side benefit is the user data captured.

Chief executive and co-founder is Deobrat Singh, who ran a software startup at school before working with Nivio, Microsoft and Electronic Arts.

The New Delhi-based team – who met at a startup event in their home city – moved to San Francisco to take part in the 500 Startups accelerator. gazeMetrix was based on a first product which allowed smartphone users to take a photo of an app on a friend’s phone and download it to their own.

An algorithm processes unique characteristics of a logo, and  uses ‘fuzzy matching’ – which copies with distorted or wrinkled images – to process social media pictures.

Images can then be aggregated, providing the opportunity to interact the smartphone users who uploaded them.