Surveillance cameras are a powerful tool used by many businesses and public facilities to prevent theft and vandalism. However, a 2008 report published by The Independent found that only 5 percent of crimes in the United States are solved using surveillance cameras. Today’s method of surveillance normally consists of security personnel tediously monitoring dozens of screens for suspicious activity. Roy-Chowdhury’s book addresses these constraints in modern camera technology, which stems from challenges in wide area tracking, the positioning of cameras, recognition methodologies and other camera design-related problems.
Roy-Chowdhury is attempting to optimize the cameras to recognize objects that may be considered suspicious. Once a suspicious object is detected, the camera system will immediately alert the security person on duty. Even if an individual manages to escape after committing a criminal act, the smarter camera system envisioned by Roy-Chowdhury might still be able to catch these perpetrators during the post-crime investigation process. Once the camera system identifies a suspicious individual, it can zoom in on the person’s face in order to facilitate the job of facial recognition software.
Andrew Bhanu, a forth-year electrical engineering student at UC Riverside believes that the research will greatly alleviate security problems. “With years of witnessing the evolution of camera technology in phones and other systems, it is great to hear about surveillance camera research that will do wonders for public safety,” stated Bhanu in an interview with the Highlander.
Roy-Chowdhury has installed 37 cameras at the Bourns College of Engineering at UC Riverside as part of his ongoing research. He believes that his book is the first to focus on camera networks in a comprehensive manner that considers the merits of image processing, computer science, mathematics and statistics.