ARTIFICIAL INTELLIGENCE VIDEO SURVEILLANCE

Views:
 
     
 

Presentation Description

Artificial Intelligence can flag people based on their clothing or behavior, detect people’s emotions, and find people who are “extraordinary”. It was the surveillance cameras that were inactive. Maybe they just recorded it, and no one else watched the video unless they needed to. The frustrated guard sees a dozen different screens, scanning for interesting ones. In both cases, the video is only stored for a few days because storage is expensive.

Comments

Presentation Transcript

slide 1:

ARTIFICIAL INTELLIGENCE VIDEO SURVEILLANCE Artificial Intelligence can flag people based on their clothing or behavior detect people’s emotions and find people who are “extraordinary”. It was the surveillance cameras that were inactive. Maybe they just recorded it and no one else watched the video unless they needed to. The frustrated guard sees a dozen different screens scanning for interesting ones. In both cases the video is only stored for a few days because storage is expensive. Growing up none of it was real. Recent developments in video analytics fueled by artificial intelligence techniques such as machine learning enable computers to view and interpret surveillance videos with human intelligence. Identity technologies make it easy to automatically identify who is in the videos. In the end the cameras themselves became cheaper ubiquitous and even better Cameras mounted on drones can effectively see the entire city. Computers can watch all videos without human problems such as distraction fatigue training or need to pay. The result was a level of intelligence that was impossible some years ago. The ACLU report published Thursday that "Dawn of Robot Surveillance" does not record ourselves as "AI-aided video surveillance" but judges us based on their perception of our actions emotions skin color clothing voice and more. Let’s take the technologies one at a time. First: video analytics. Computers are improving in detecting what is happening in the video. It is easy to identify when a person or vehicle enters a prohibited area. Modern systems become alert when someone is walking in the wrong direction — for example an exit only passing through the corridor. They can count people or cars. They can identify when the luggage is left unattended or when the luggage is taken away. They can detect when someone is breaking into an area lying down or running. Increasingly they can identify the specific actions of individuals. Amazons cashier-less stores rely on video analytics to figure out when someone has taken an item off the shelf and not put it back on. Slightly recognizing actions video analytics enable the computer to understand what is happening in the video: they can flag people based on their clothing or behavior identify peoples emotions by body language and behavior and find people who are "abnormal" on a per-person basis. Around them. The same Amazon store cameras can analyze customer sentiment. Other systems can explain what is happening in the video scene.

slide 2:

Computers can also detect people. AIs are improving in identifying the people in those videos. Facial recognition technology is constantly improving facilitating the enormous storage of tagged photographs we provide to Facebook and other social media sites as well as government-issued photos in the process of issuing ID cards and drivers licenses. The technology is already there to automatically detect everyone who “sees” the camera in real-time. Even without video recognition we can find unique information that is constantly transmitted through smartphones that carry with us everywhere or through our laptops or Bluetooth-connected devices. Police have been tracking phones for years and this method can now be combined with video analytics 5 reasons AI needs to evolve video surveillance solutions Why video surveillance solutions Artificial intelligence has begun to impact the industries on a large scale but the impact is nowhere near as high as the security column. New and innovative solutions are being launched not only by establishing security vendors but also by smaller startups and these solutions add value to a companys security operations. Let’s discuss how AI will impact the video surveillance industry shortly.  Real-time monitoring Basically in the days of CCTV cameras a video was used to live on TV screens but very little was done to make any meaningful analysis of the security incident. In those days video surveillance solutions were always reactive and continued to be used in large parts of the world. Most agencies for CCTV footage are only in the event of an incident or a massive threat awareness. The use of AI has completely transformed the picture through real-time monitoring of video footage with a clever analysis of how security incidents can occur and what needs to be done to prevent them from happening. Various large IT vendors have come up with AI solutions that can detect unattended items such as bags at airports and public places allowing security agencies to take action before it is too late. Similarly AI-enabled surveillance solutions can monitor people rotating in retail stores and identify who is involved in a theft.  More accurate than men CCTV monitoring is as accurate and reliable as the monitor on the screen.However being human operators sometimes lose crucial security credentials which can have devastating consequences for the company. Fatigue is one of the reasons why people dont regularly watch video footage. The advantage of AI-based solutions is that they can be trusted that machines are accurate at all times and can address human error. This ensures that no details no information and absolutely no threat is detected and therefore the companys security is assured.

slide 3:

Foolproof deployment of technology such as an object event and face recognition AI technologies bring superior facial recognition object recognition and event recognition capabilities thus providing active and real-time security. This is a huge application for the police force because they need to identify criminals from the crowd in large areas. Another interesting application is the marketing departments of large corporations where they need to identify high-net-worth customers from their stores to provide them with high-value services. Another important way of identifying a person through faceless identification is where a persons physical features such as height posture and structure can be used to identify them in a group. Also patterns of activity can be used in different environments to isolate criminal behavior and keep neighborhood neighborhoods safe.  Image processing allows for better analysis Although ultra-high-definition cameras are capable of capturing high-resolution images their use as part of video surveillance solutions is limited. As a result most of the images captured by such solutions are of poor quality especially in low light conditions. This can serve as a barrier when doing meaningful analysis with such video footage. Artificial intelligence can help security personnel sharpen low-quality images to extract meaningful information from them. This helps to make good use of obscure images and videos for useful analysis of the eye.

slide 4:

 Making better use of the data flood There is a web of cameras that capture valuable security footage in key geographical locations. Though legacy solutions did nt allow much to be done in terms of meaningful analysis of such big data. Special software based on the latest technology is required to work on this massive data and create security alerts based on comprehensive analysis. This enables security officials to understand big data in a qualitative manner and take crucial steps to protect human beings and valuable assets. P.Venkat Vajradhar Marketing TeamSEO Specialist USM SYSTEMS 8-2-293/82/A/270E Road No – 10 Jubilee Hills Hyderabad-500034.

authorStream Live Help