The Internet has gone through several seismic changes in capability and content over the years. Successes have included protocols for electronic mail, cross-linked Webpage documents, search engines that are still almost entirely based on analysis of text, and formats for sharing print-quality documents (PDF) and audio and visual media (MPEG). The emergence of cheap cellphone cameras and webcams has provided a significant portion of that media, which people have chosen to share in vast quantities. Facebook and YouTube could not have emerged as they did without these cheap networked cameras.
Another massive game-change is coming, and there will be huge opportunities to organize and capitalize upon this change. Cameras will develop a few new capabilities, including real-time streaming through various protocols such as 4G and advanced WiFi, embedded computer vision processing, back-end computer vision processing on servers or across a network, and subscription-based sharing services that allow an individual to benefit from all the other individuals who choose to share their data.
Another very important development will be the emergence of cameras in vehicles, and cameras worn by individuals. Streaming implies real-time relevance, and mobile cameras can provide qualitatively different information than cameras bolted to lamp-posts or buildings. Already, many fleet vehicles (delivery trucks, taxicabs, police cruisers) have at least two cameras, one facing out the front and one into the interior, for insurance and security purposes. Insurance companies will inevitably encourage such devices in personal vehicles. There is also a latent market for civilian 'personal surveillance' devices worn on the lapel for personal security, based on devices currently worn by some law enforcement and military personnel. Legal battles are already underway in many U.S. states to determine how such imagery can be used.
What will make these rapidly proliferating devices most useful are the computer vision methods that are emerging today to exploit the imagery and make collective sense of it. We explore these methods in part 2 of this post and discuss what new information we can gain from combining this new technology in part 3.
The author's affiliation with The MITRE Corporation is provided for identification purposes only, and is not intended to convey or imply MITRE's concurrence with, or support for, the positions, opinions, or viewpoints expressed by the author.