The brain of the video surveillance camera is the chipsets that integrate the functionality of complex megapixel cameras. Traditional chipset providers, such as Ambarella and Hisilicon, are able to deliver their customers megapixel and sophisticated camera functionality in silicon today.
Intel recently introduced their new, faster Atom E3900 series which has more powerful graphics and image processing. The E3900s are designed for a wide range of applications, including manufacturing and surveillance, and they’ll soon be joined by a version specifically for vehicles, called the A3900.
Intel is working to help machines evolve from accurately sensing what’s going on around them to acting on those senses. For example, if a device can see defective parts going through an assembly line, it can alert someone or even stop the line. Cameras in cars could see that the driver is drowsy and set off an alarm in the car, and ones pointed in front of the vehicle could tell a pedestrian from a shadow and stop the car. Intel has also produced a software platform to carry out video analytics.
At the heart of Qualcomm’s new connected-camera push is a new processor, the Snapdragon 625, that boasts 25 percent lower power consumption than earlier chipsets. It has the imaging muscle to capture 24-megapixel images and 4K video at 30 frames per second, but also the capability for deep learning, Qualcomm says. The chip also has a reference design to help manufacturers get products on sale more quickly. It expects the design to reach some makers by the end of 2016, with cameras based on it hitting the market soon after.
Qualcomm is also rolling out a video analytics API and connected camera SDK. They bring image processing and analytics software capabilities including voice activation, face detection, recognition and object tracking to the camera. These features can make cameras smarter about what to record so there’s less content to store, transmit across a network or sift through later. A surveillance system may be less expensive to operate if the camera can tell what events are worth watching and only turns itself on when it detects something interesting.
Qualcomm is also adding Linux support to its connected-camera lineup, on top of Android, which is already there. This will give product developers a broader software ecosystem and more security options.
In September last year Intel acquired Movidius a fabless semiconductor company. They design their own chips, but outsource the actual manufacturing of the chips, but given Intel's chip fabrication capabilities this will no doubt change in the near future.
Since the acquisition Movidius has signed partnership deals with Hikvision and Flir. Hikvision will soon be incorporating Movidius Myriad 2 Processing unit a system on a Chip (Soc) device.These devices integrate deep learning and this should increase the accuracy and performance of video analytics. Nvidia’s has deep learning server based analytics and some edge based chips but Movidius chips are a fraction of their price.
There has been a gap in the market for some years to deliver improved analytics in video cameras and if Hikvision can meet this demand through Movidius they will strengthen their grip on the market. However Qualcomm, Ambarella and Hisilicon will all be offering improved SoC devices.
What is clear is that Big Chip manufacturers are now Big IoT developers. They can deliver an end to end solution by producing new camera chips and communications chips plus software platforms that can operate on the cloud. The battle on who wins the major share of the IoT Business is about to start.
This article was taken from Memoori’s Eighth edition of their Annual Report The Physical Security Business 2016 to 2021; which can be purchased online for $1,500 USD for a single user license.
[contact-form-7 id="3204" title="memoori-newsletter"]