Human Interaction Proofs (HIPs) based on Emerging Images and Topological Data Analysis (TDA) Techniques

Abstract

In this paper, we introduce an algorithm to generate emerging images which can be used to implement Human Interaction Proofs (HIPs) based on this kind of images. In a general setting we propose the design of proofs easy to solve by human beings but hard to solve by computer algorithms. These tests are able to tell apart human from bots (or robots) in order to protect online environments (as social networks, wikis, ticket sellers, free-email providers, etc) from different kind of security threats.Our proposal uses tools and techniques arising from TDA (point clouds, simplicial complexes and spatial triangulation among others), in order to create shapes which can be identified by humans as recognizable images hard to detect by machines.