Smart video surveillance using artificial intelligence

One of the biggest challenges in brick-and-mortar retail is loss prevention to deter shoplifting. However, label fraud, customers unintentionally failing to pay, and errors made by checkout staff also result in financial losses for retailers. Smart video surveillance systems using artificial intelligence can make a significant contribution to reducing stock loss, particularly in an era of open entrance areas without access barriers and self-service checkouts. Furthermore, the analysis of customer behaviour through intelligent video surveillance and video analytics holds great potential for optimising product ranges. The following practical examples illustrate the current situation and offer a glimpse of future developments.

Identifying shoplifters

One key area of application for smart video surveillance is the identification of shoplifters. This is because shoplifters exhibit a typical pattern of behaviour that can be detected by algorithms in real time. To achieve this, the software must first be trained to distinguish between desirable and undesirable behaviour. Typical behaviour of shoplifters includes, for example, constantly looking around, searching for CCTV cameras or lingering in one spot for a long time. Whilst a security guard can also recognise these suspicious behaviour patterns, the AI-based algorithms of a video surveillance system analyse hundreds of customers simultaneously in real time.

Detect items in the shopping cart

Normally, it has been the job of checkout staff to take a look inside the shopping trolley. Depending on the situation at the checkout, this process can take a few seconds, thereby slowing down the checkout process and reducing attention to the rest of the shop floor. AI-based video surveillance with monitoring cameras for shopping trolleys, on the other hand, independently determines whether there are still goods in the trolley. If this is the case, an alert message is displayed on the till monitor so that the checkout staff can respond accordingly.

AI-based analysis of shopping basket contents also has practical applications in other situations. For example, a video surveillance system can detect whether customers are leaving a shop with a full shopping basket via the open entrance area, in breach of the rules. A member of staff is then notified in real time to investigate the situation. Furthermore, intelligent video surveillance using artificial intelligence can analyse the contents of shopping trolleys in the queue and use this data to recommend opening additional checkouts, thereby reducing waiting times for customers in the queue.

Check the self-service checkouts

 Although the number of self-service checkout areas is growing only gradually, they are becoming increasingly popular with customers. To keep stock losses to a minimum, smart video surveillance using artificial intelligence significantly reduces both unintentional operational errors and deliberate theft. This is because the video surveillance system uses AI-based algorithms to intelligently analyse transactions at the self-service checkout in real time. If any irregularities are detected, checkout staff are automatically notified.

Analysing shopping behaviour

​

Smart video surveillance is not only intended for loss prevention, but can also help analyse customers’ shopping behaviour. This is because video cameras are now capable of identifying visitors’ gender and age. Combined with analyses of customer pathways and dwell times, it is possible to determine which customer groups frequent certain routes more or less frequently. These insights then enable optimisations to be made to the product range, product presentation and staff deployment. AI-based video analysis can also specifically examine how customers engage with certain products and their purchasing decisions, and derive recommendations for action from this.

Get started with smart video surveillance today

Smart video surveillance and real-time video analysis, particularly when dealing with large volumes of data, are only just beginning to realise their full potential. They are a key component in reducing stock losses, supporting innovation in the retail sector, and gaining an even better understanding of customers’ shopping behaviour. We would be happy to advise you, drawing on our expertise in the common applications of smart video surveillance using artificial intelligence, and design an AI-based video analytics system tailored specifically to your needs.