Success to BI, not to AI predicted in 2020

However, for the security CCTV industry AI still means ‘advanced analytics’, not ‘artificial intelligence’, as the term artificial intelligence literally means computers’ ability to think. CCTV analytics allows recognising objects, tracking and classifying them, but it is still way far from providing any kind of assessment of the situation based on revealed details and automatically execute any actions, replacing live operator.

The BitRefine group believe that people will manage to combine various technologies such as machine learning, reinforcement learning, generative adversarial networks, augmented data learning. All of this will ultimately bring us to fully automated surveillance, but it is not going to happen in 2020 for sure.

For 2020 BitRefine group sees the greatest opportunity in turning a CCTV system into a source of business intelligence (BI). Big companies are already exploring approaches that will allow them leveraging data, generated by existing video surveillance to improve three general areas: increase yields, reduce costs and improve competitiveness. The actual range of possible use cases is limited only by level of understanding of company’s business processes and available data. According to some estimates, in terms of improving business efficiency proper use of existing video assets bring from one to ten percent of cost savings.

Surveillance video is a typical example of big data. And besides, it is extremely rich ‘big data’. The questions is what kind of insights a company wants to extract from it. Today advanced video processing systems, based on deep neural networks can automatically extract almost all kind of information from real-world footages. There’re already corresponding commercial tools on the market. 

Now it is business’ turn to see the value of the data that their CCTV systems possesses and to apply business intelligence to increase the efficiency of physical processes.