Qualcomm adds smart camera SoC targeting surveillance in smart cities - Embedded.com

Qualcomm adds smart camera SoC targeting surveillance in smart cities

New Qualcomm QCS7230 processor is aimed at enterprise security and public safety, using connected intelligent edge cameras.

Qualcomm Technologies has introduced a new smart camera system on chip (SoC) as part of its Vision Intelligence Platform portfolio to enable new intelligent scalable surveillance capabilities in smart cities, spaces and enterprises.

Qualcomm QCS7230 chip
(Image: Qualcomm)

The new Qualcomm QCS7230 processor is aimed at enterprise security and public safety, using connected intelligent edge cameras. In a briefing to announce the new camera solution for the International Security Conference and Exposition (ISC) West, Siddhartha Franco, director, business development, Qualcomm Technologies, explained the big shift in business models for surveillance. Traditionally, industry has been using video management software installed at one time on site and with manual monitoring and fixed analytics, with very little capability for remote access or management and upgrades.

But that is changing to video surveillance as a service, with long term dynamically upgradeable services which can scale easily with new AI and ML models and as requirements and trends change; these can be trained in the cloud or edge, and analytics can then be performed in real time at the edge.

The focus of the new camera solution is to address this shifting trend, featuring the connectivity to support expanded end-to-end solutions for smart cameras through the IoT-as-a-service (IoTaaS) model. It also supports customers looking to transition from traditional video management software (VMS) models to end-to-end service-based capabilities with IoTaaS.  It also helps accelerate edge AI services with the building blocks needed to support businesses and entities looking to deploy smart cameras and intelligent IoT devices to support video collaboration, access control, enterprise and home security, 360-cameras, dash cameras, and wearable cameras.

qualcomm-qcs7230-block diagram
The QCS7230 block diagram. (Image: Qualcomm)

In terms of its position in Qualcomm’s portfolio, the QCS7230 sits just below the premium QCS8250 which the company introduced in June 2021. The QCS7230 offers a heterogeneous computing architecture coupled with the Qualcomm AI engine to efficiently run complex AI and deep learning workloads and on-device edge inferencing at low power. The processor also offers an image signal processor (ISP) with support for up to seven concurrent cameras, or up to 24 video streaming cameras, a dedicated computer vision engine for enhanced video analytics (EVA), as well as the Qualcomm Hexagon Tensor Accelerator (HTA). It also supports Wi-Fi 6 connectivity speeds via a companion module, making it suitable for developing IoT devices and enterprise grade solutions.

Franco said, “With the digital transformation of industries there has been a heightened focus and need for the growing number of smart devices at the connected intelligent edge. Today’s cameras require superior connectivity and interoperability, powerful edge computing and AI capabilities in addition to protections for security and privacy.” He added that Qualcomm Technologies’ portfolio of smart camera solutions make it easier for businesses and entities looking to deploy smart cameras to support a variety of use cases across industries, from creating safer communities to informing better business decisions.

Security industry increasingly needing help with video analytics

Security and surveillance have seen an explosion both in deployment, especially during the pandemic, as well as in image quality and volume. With security departments having to see and analyze video and data securely and efficiently to provide meaningful interpretations of a scene, smart cameras equipped with connectivity, AI, and analytics can help address the need to better handle this deluge of images. As a result, this helps support cities, businesses, and entities with the real-time intelligence needed for enhanced insights that can improve the quality of life for communities and drive enhanced business decisions and outcomes.

Qualcomm said there are three key trends driving innovation in smart cameras to move the industry beyond the initial wave of cameras using computer vision and enable the security industry to take a giant leap forward.

The first is that cameras are beginning to see more because sensors are rapidly transitioning from full HD (FHD) to 4K resolutions, plus camera makers are adding multiple sensors to a single camera unit, enabling higher utilization of the underlying hardware.

Secondly, compute capabilities on cameras are rapidly increasing, and companies are innovating with newer machine learning models in response to the high demand of a variety of use cases, such as people and object detection, classification, people counting, and vehicle recognition.

Third, the ubiquity of security cameras makes it possible to install cameras even in hard-to-wire areas. In particular, with 5G becoming more pervasive, the low latency and high bandwidth of a 5G connection enables new use cases, including mission critical emergency services, smart factories, and smart cities.

Qualcomm IoT Smart Camera ecosystem
In order to help deployment of smart camera solutions, Qualcomm has also developed an ecosystem of partners. (Image: Qualcomm)

In order to help deployment of smart camera solutions, Qualcomm has also developed an ecosystem of partners. For example, Thundercomm has introduced a commercially ready TurboX QCS7230 system on module (SOM) to enable developers and manufacturers to quickly develop smart cameras and intelligent IoT devices across multiple applications. The TurboX QCS7230 SOM is available now and enables simple, security-rich and high-performance camera connection and AI analysis.


Related Content:

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.