When Do Video Systems Yield Information Other Than Images?

Retail analytics can effectively transform a network camera into a very powerful marketing tool that enables retailers to find out more about their customers and improve the shopping experience. By analyzing the objects within the video image, we can determine how many people are coming in or out of a location, how many people are inside the location, the average visit time per customer, line levels, and identify customer demographics by age and gender. This information, combined with basic sales data, enables the retailer to make intelligent business decisions and to optimize operations, increase sales, and offer better customer service. Some retail analytics perform all image processing at the edge or completely embedded within the network camera. Therefore, no images or video are collected to be analyzed later. Data is completely anonymous, effectively converting the camera into a sensor while still providing the benefits of a network video solution.

Video analytics and the resulting metadata is being analyzed to provide solutions that go outside of security and even video surveillance. The cameras are being used as smart sensors in an ever-growing number of applications. For example, thermographic cameras are widely used in auto safety applications, such as animal and pedestrian detection, as well as in industrial applications for leak detection or high temperature warnings. Retail marketing widely leverages digital video analytics in applications such as people counting, heat mapping, traffic patterns, and analyzing line times. Machine vision is using digital video for vision-controlled robotic handling systems used in applications such as product packaging, and automated harvesting. The information derived from digital video images assists in an ever-growing number of applications.

In today’s competitive business environment, organizations of all types and sizes seek new ways to transform video and related data into valuable insight — propelling more informed, intelligent decision-making. This process allows stakeholders to make sense of video data to detect risk proactively and to activate efficient response. Transforming raw data into actionable intelligence helps increase efficiency, enhances safety and optimizes business practices. Examining video and understanding its correlation with other systems and processes delivers a broad range of benefits — it allows organizations to uncover hidden patterns, unknown correlations, market trends, customer preferences and other useful business information. Stakeholders can gain insights that empower them to anticipate, respond, and take action — which are necessary to successful operations. With the amount of video growing rapidly, the process of understanding how it affects ongoing operations is critical. Businesses that generate actionable intelligence from video are better positioned to achieve strategic objectives.

Video surveillance is a critical addition to data analysis and situational awareness. For example, in a city environment, first responders can leverage video to assess a situation in real time and execute a more informed and quicker response to a situation. Casino surveillance operators can use cameras to monitor for compliance with industry and government regulations. Oil and gas facilities can monitor system operations in remote sites to prevent system failures. Marketing teams also find video useful to determine the success of promotional campaigns within retail stores. These scenarios point to the growing benefit of video surveillance as a security and business optimization tool. But as the value of video increases, users need to ensure this investment is protected. Quite simply, downtime and data loss is not an option for any business, whether it operates one site locally or many diverse facilities across the globe.

Data on the identification of people and other objects for the purposes of counting them, tracking their movements or identifying the, are useful for a number of applications such as customer analytics in retail, tracking of vehicles for traffic management, and identifying particular individuals or objects for security or other purposes. The "value" extracted from the data can be realised in various ways; for example, customer analytics in retail can be used to increase sales by making marketing more effective or optimising store layout according to traffic patterns within a store. The accurate identification of people by facial recognition could be valuable for applications such as restricting access based on identity or providing differentiated customer service in retail or hospitality. The automatic recognition of automobile number plates is valuable for applications such as parking management and congestion pricing.