Video analytics: Cutting Edge AI

• Jan 5, 2023
• Grant Furlane, Locomobi World

As we step further into the age of Artificial Intelligence, one theme that stands out is the merging of video and machine learning to better surveil our world. The term Video Analytics best describes the multiple ways in which this field has revolutionized the use of cameras and data to improve upon human intelligence.

The idea of video analytics is integral to the concept of Smart Cities. These are the idealized cities of the future that will be built to afford the greatest comfort, space, and intelligent design for the resident humans with the help of AI. Analytics will be used to make a community happier, with better health care and business support. There will be thoughtful planning on where to locate green spaces and create energy-efficient homes. The Smart City is the blueprint for future city construction.

Video analytics systems have been programmed to mimic human behavior while leaving out human error. What is known as Deep Neural Networks trains the system to identify specific objects and track them..

Lizzie Goldmeier, Director of Marketing at BriefCam, in her podcast explains the finer points of video analytics, describing the advantages and usages of the concept. Infinite man-hours can be saved, problems detected in minutes, and all backed up with data. As she explains, “it takes video and transforms it into something easy to analyze and use”.

The public is mainly unaware of how pervasive video analytics is in this modern world. There are several aspects of everyday life where it is employed constantly. Traffic, not just by auto although that too, but foot traffic in supermarkets and hospitals, bathroom traffic, entrance, and exit traffic, all are scrutinized. The results influence decisions on everything from employee hiring to road safety regulations.

Health Care is one area of increased use of video analytics. Hospitals use surveillance to oversee staff, patients, and visitors. Theft of drugs and infant abduction are ongoing problems where increased surveillance is enforced. How often patients are checked, and appropriate drugs administered are easily detected as are visitor traffic, wait times, and accessibility of emergency areas.

Of special importance is at-home monitoring. Once away from the surveillance of the hospital environment, analytics can monitor patients, especially the elderly, to assure there are no falls. Also, the proper time and dosage of medication can be detected, and it will be known that it is taken.

Not everyone is thrilled by the idea of being always under surveillance. Not only physical issues but mental healthcare can be addressed by analytics. Facial expressions, body posture, and even intensity of gaze can all be used in evaluating the patient. The subtlest of expressions that would be missed by the human eye will be picked up on by video analytics. Children with autism spectrum disorder can be identified with the use of a simple phone app that can track facial expressions and eye movements.

The police were quick to pick up on the crime-fighting benefits of video analytics. From license plate retrieval to traffic violations the video doesn’t lie. When searching for a specific car, the analytics can scan in a matter of minutes footage that it would take enormous man hours to sort through. Facial recognition software can place someone at a crime scene or ATM machine without a doubt.

Traffic patterns can help determine city planning and identify where problem areas exist. If there is a bottleneck area of congestion, it can be attended to. Operational decisions are viable when they can be backed up with data. Entrances and exits can be kept open and the traffic kept flowing. In some countries, a speeding ticket has become a thing of the past. People receive a fine in the mail for having been caught on camera going over the limit.

With the increase in urban traffic, increased traffic jams and accidents are inevitable. Video analysis plays an important part in traffic management. Traffic analysis can adjust traffic light control and study why congestion happens in certain areas. It can detect accidents in real-time and offer proof if a case goes to court.

The right to privacy is at odds with the need to regulate a congested world. Although companies involved in AI may have their own standards and are encouraged to act ethically, the fact is that there are no laws legislating the right to privacy in the public domain. Any time you use a social media site you are publicizing yourself to the world. Universal surveillance is a part of the human experience now.