Perhaps the biggest technology trend in security over the past several years has been the evolution of artificial intelligence (AI) and the enhanced capabilities offered to end-users through the development of machine/deep learning algorithms. The potential applications for this technology have also subsequently opened the door for a wave of new market entrants in both the emerging field of security robotics and the more traditional video surveillance arena.
One company that falls into this latter category is Umbo Computer Vision, a developer of autonomous video security solutions. Though the company has only been around for about three-and-a-half years, Umbo has already made an impression on many within the industry and was even recognized with the “Best in Video Analytics” award for its’ Umbo Light solution during the SIA New Products Showcase at ISC West 2017.
According to Shawn Guan, Umbo’s CEO, the company is focused on providing end-to-end solutions that can deliver real-time event detection in the most frictionless way possible for end-users. “We can actually recognize the behavior of people in the video from using a single image. The result is really accurate event detection for customers,” Guan says.
Late last month, Umbo launched its latest camera, the Umbo SmartBullet, to complement their existing SmartDome camera. The Umbo SmartBullet features a motorized lens and is three times more powerful than the SmartDome, enabling them to send a 1080p video stream at 60 fps or two 1080p streams at 30 fps. Both the SmartDome and SmartBullet feature 30 days of onboard storage through the use of an SD card and are controlled and managed via the cloud, thus eliminating the need for an NVR.
Guan says the SmartBullet also features a modular design which gives end-users the ability to add various accessories to the camera to better suit their surveillance needs should their requirements change over time. Among these accessories include:
- An extra IR piece that can be attached to increase the IR distance up to 160 feet;
- An LED “torch light” to deter would-be intruders or aid in the detection of colors and other details that would have otherwise remained hidden;
- And, a long distance microphone to capture audio where necessary.
According to Guan, the main difference between the intelligent video solutions offered by Umbo and some of the others on the market is that they don’t rely on some of the traditional methodologies employed by previous generations of video analytics.
“We have our own camera, cloud (service) and, at the same time, we have our algorithm for event detection. And our event detection is video-based instead of image based,” Guan explains. “Some companies that develop machine intelligence might do facial recognition but a lot of these technologies are based on frame-by-frame (analysis), which means they look at that single image and compare that to see if it is the same person. We focus more on behavior, so that is what really separates us.”
For example, Guan says if someone were to attempt to scale a wall in an area covered by one of their cameras, their technology would identify that behavior and would not be reliant on someone entering a predetermined box or line within a frame, which is the most common way video analytics were setup to perform in the past.
Guan says that Umbo has gone to market exclusively through the integrator channel and that the majority of their deployments to this point have been within the commercial sector, including small and medium-sized businesses (SMBs) and even some Fortune 500 enterprises. Their cameras have also been used within the education sector in the U.S. and for some public safety deployments internationally.
About the Author:
Joel Griffin is the Editor-in-Chief of SecurityInfoWatch.com and a veteran security journalist. You can reach him at [email protected].
