Years of double-digit growth in shipments of video surveillance cameras in China are over, giving way to future growth in deep learning video analytics.
Years of double-digit growth in shipments of video surveillance cameras in China are over, giving way to future growth in deep learning video analytics. According to the latest Video Surveillance Intelligence Service by IHS Markit, unit shipments of video surveillance cameras grew just 2.3% in China in 2016 compared to the previous year, at 58.2 million units, much lower than in 2015 (34.6%), 2014 (38.5%) and 2013 (29.6%). The lower rate of growth in security camera reflects the high installed-base of cameras in China. Demand for cameras in future will be increasingly due to replacements rather than new installations.
However, Chinese video surveillance market revenues are still forecast to grow at a compound annual growth rate (CAGR) of 12.4% between 2016 and 2021 from $6.4billion to $11.4 billion. The major driver is deep learning for video analytics. And much of the future growth is forecast to come from recorders rather than cameras. The recorder market will be driven by demand for deep learning-based video analytics servers and video management platforms with integrated video analytics. This also explains why the average price of a recorder in China is forecast to rise.
With numerous large projects of surveillance camera installation and massive amounts of video data, end-users in China are eagerly seeking ways to interpret this data with video analytics. Examples of such end-users include police and traffic departments.
Deep learning is poised to revolutionize the video surveillance industry and facilitate a leap in the capabilities of video analytics.
Deep learning is poised to revolutionize the video surveillance industry and facilitate a leap in the capabilities of video analytics. Graphics processing units (GPUs) are currently the most common solution employed by video surveillance vendors to run video analytic applications. Facial and vehicle recognition, for example, are two applications that vendors claim have benefited from applying deep learning techniques. Not only does deep learning increase the accuracy of facial and vehicle recognition sensors, they also claim it enables faces and vehicles to be identified in larger and more crowded scenes.
In 2017, facial recognition has started to become a requirement in new city surveillance projects in China. It is often specified in tender documents for safe city projects, signaling that the shift toward deep learning-based video analytics is well underway in China.
There is also high potential for deep learning-based video analytics to be used in the Chinese retail sector to enhance business efficiency and improve customer experience. “No-checkout” retail stores are one example where this kind of technology could be used. These stores allow customers to walk in, select products and walk out again, without them needing to stand in a checkout line. Payment is made automatically. Amazon launched a “no-checkout” store called, Amazon Go, in late 2016. Alibaba Group also recently opened an experimental cashless supermarket in China, called Tao Café. Tao Café uses face recognition and sensor fusion technologies to provide its no-checkout experience.
Deep learning is an emerging technology in the video surveillance industry. However, demand in China will grow quickly, providing the engine for future market growth.