A world filled with cameras watching streets, stores, buildings, and hallways brings to mind two diverging images. One is of an ability to catch crime or detect emergencies as they happen, increasing security and safety. The other, however, is an image of some faceless watcher spying on everything everyone does, eroding personal privacy.
Both images have proven to be correct. Video analysis systems routinely alert security personnel to theft and intrusion, and surveillance cameras often prove instrumental in identifying and convicting perpetrators. These same video systems, though, have also been used for blackmail, voyeurism, and discrimination.
The Swiss company Emitall Surveillance has addressed this conflict with technology that scrambles the images of identified objects to protect privacy. Thus, a video analysis system might detect people walking down a hallway, but the Emitall technology would scramble their images so they are unrecognizable. This allows the system to still detect an intrusion into a restricted area without allowing system users to identify people merely passing by.
If identification is necessary, such as finding out who broke into the building, Emitall’s scrambling technique is reversible. The system encrypts the data when scrambling it so a high-level authority such as the courts can control the decoding key. This approach protects privacy while retaining all the security benefits video analysis offers.
Reprints
