Security Applications for Embedded Vision
Vision security, or intelligent video, provides perimeter intrusion detection
Vision products have been used in the physical security market for a number of years. The technology is commonly referred to as video content analysis (VCA) or intelligent video in the physical security industry. It is used in virtual tripwire applications, providing perimeter intrusion detection for high-risk facilities such as airports and critical infrastructure.
Typical projects use video analytics to generate real-time alerts when an object is identified and tracked through a pre-defined area. Operators, based in the central control room, can manage more video streams as they are not required to identify intruders purely from watching the monitors. Instead, the vision product will identify and alert the operator when there is a potential security issue. It is then up to the human operator to decide how best to deal with the threat.
Other applications of vision products in the security space include wrong-way detection and trajectory tracking. These types of algorithm can be used in airport exit lanes, where security managers want to identify people entering the restricted area from the main terminal. This responsibility has historically been given to security guards; however, vision technology can reduce security guard costs for airport operators and increase the accuracy of the security system.
Other applications in the security market include behavior recognition algorithms such as “slip and fall” and abandoned/left object detection. These applications are less developed than the perimeter detection systems, but will provide powerful tools for security managers over the coming years.
What are the primary vision products used in security?
The primary vision products in the physical security market are network cameras, encoders, NVRs (network video recorders), DVRs (digital video recorders) and intelligent appliances. Established video surveillance vendors supply the majority of these products. These vendors either develop analytics in-house or embed third party algorithms from video content analysis software developers. Consequently, growth in the market is heavily driven by what these vendors decide to do and the extent to which they commit to intelligent video.
Over the last couple of years, the video surveillance market has also seen “low-end” analytics, such as camera tamper detection and video motion detection, become a standard feature on many vendors’ product ranges. These applications have developed from a distinct video analytics market to a product feature used as a differentiator. “Low-end” analytics are embedded on cameras, encoders and storage devices.
Fisheye dewarping with no need to install an app: this solution works directly with the WebGL API and on popular browsers for computers and mobile devices, as long as they support HTML5. Convenience and ease-of-use should be watchwords for the development of all products, yet many of them stumble right out of the blocks by
“Multi-modal Re-identification: IOT + Computer Vision for Residential Community Tracking,” a Presentation from Seedland
Kit Thambiratnam, General Manager of the Seedland AI Center, presents the “Multi-modal Re-identification: IOT + Computer Vision for Residential Community Tracking” tutorial at the September 2020 Embedded Vision Summit. The recent COVID-19 outbreak necessitated monitoring in communities such as tracking of quarantined residents and tracking of close-contact interactions with sick… “Multi-modal Re-identification: IOT + Computer
This blog post was originally published at Lattice Semiconductor’s website. It is reprinted here with the permission of Lattice Semiconductor. If you were to ask them (and I have), you would discover that many people think of radar in the context of things like airplanes and ships and the evening weather forecast on TV. As
This blog post was originally published at Ambarella’s website. It is reprinted here with the permission of Ambarella. Advances in our HDR processing algorithms are changing the way cameras handle challenging lighting conditions. In everyday life we often encounter strong lighting contrasts—shadows under the sun, bright lamps at night, a garage door opening to reveal
What’s New: Today, Intel introduced Intel® RealSense™ ID, an on-device solution that combines an active depth sensor with a specialized neural network designed to deliver secure, accurate and user-aware facial authentication. Intel RealSense ID works with smart locks, access control, point-of-sale, ATMs, kiosks and more. “Intel RealSense ID combines purpose-built hardware and software with a dedicated
“Video Activity Recognition with Limited Data for Smart Home Applications,” a Presentation from Comcast
Hongcheng Wang, Director of Technical Research at Comcast, presents the “Video Activity Recognition with Limited Data for Smart Home Applications” tutorial at the September 2020 Embedded Vision Summit. Comcast’s Xfinity Home connects millions of home smart cameras and IoT devices to improve its customers’ safety and security. The company’s teams use computer vision and deep
This blog post was originally published at Ambarella’s website. It is reprinted here with the permission of Ambarella. In an age of universal surveillance, our sensing camera solutions offer a privacy-focused alternative. Security cameras are seemingly everywhere, all around us. They’ve become so ubiquitous that we barely even notice them anymore, monitoring and recording us
OUTLINE: Market figures & trends: 2020 has been exceptional for thermal imaging and sensing technologies. The thermal imaging market will reach around US $7.5 billion market value by 2025, at an 8% CAGR from 2019 to 2025. The thermal detector market will jump in 2020 by 23% in value to reach US$362 million. Smart building
This blog post was originally published at Ambarella’s website. It is reprinted here with the permission of Ambarella. The next generation of access control is being shaped by both security and health concerns. Ensuring safe access to buildings and rooms has arguably never been more critical. Or, unfortunately, more complex. What was once purely a
OUTLINES: Market status: The total 2019 surveillance camera market exceeded US$20 billion. Effects of the trade war and the coronavirus were minimal. China: the surveillance camera market is exploding. The processing market is very fragmented with many different players, particularly in China, and it will remain stable. Regarding the computing market, a Sino-American duel has
To Catch a Poacher: How Our Engineers Brought AI Tech to the Fight Against the Illegal Wildlife Trade
This blog post was originally published at Intel’s website. It is reprinted here with the permission of Intel. In the wildlife reserves of East Africa, elephants, rhinos, gorillas, and other large mammals are hunted by poachers. All that stands between these animals and harm’s way are small teams of park rangers and conservationists. The danger
This article was originally published at Tryolabs’ website. It is reprinted here with the permission of Tryolabs. In the new world of coronavirus, multidisciplinary efforts have been organized to slow the spread of the pandemic. The AI community has also been a part of these endeavors. In particular, developments for monitoring social distancing or identifying
Video Analytics Solutions Are Driven by the Need for Safety and Security, With Global Revenue Reaching $4.5 Billion in 2025
The Rise of Artificial Intelligence Has Been Instrumental in Drastically Raising Accuracy Levels to Meet Market Demands The world of computer vision has evolved rapidly in recent years, and computers are now more accurate than humans in some application areas. A new application of computer vision called video analytics (VA) has emerged. VA deals with
Ilya Brailovskiy, Principal Engineer, and Changsoo Jeong, Head of Algorithm, both of Ring, present the "Optimizing SSD Object Detection for Low-power Devices" tutorial at the May 2019 Embedded Vision Summit. In this talk, Brailovskiy and Jeong discuss how Ring designs smart home video cameras to make neighborhoods safer. In particular, they focus on three key
William “Bill” Santana Li, Co-founder, Chairman and CEO of Knightscope, talks with Vin Ratford, Executive Director of the Embedded Vision Alliance, for the “Visual AI Enables Autonomous Security” interview at the May 2019 Embedded Vision Summit. Knightscope, a physical security technologies company based in Silicon Valley, develops and sells a line of autonomous robots that
The Embedded Vision Summit was held on May 20-23, 2019 in Santa Clara, California, as an educational forum for product creators interested in incorporating visual intelligence into electronic systems and software. The presentations delivered at the Summit are listed below. All of the slides from these presentations are included in… May 2019 Embedded Vision Summit