Question:

Greetings, fellow video system experts. I am seeking your advice on the current state and capabilities of video analytics software for retail security. I work for a large optical retail chain in Europe, with over 300 stores across the continent. We are facing a serious problem of internal theft and loss prevention, especially after the pandemic. Our previous method of manually reviewing video footage was very time-consuming and inefficient.

I have suggested to my manager that we should consider using some video analytics software to automate and enhance our security system. I have done some preliminary research on the available options, such as BriefCam and Pelco , but I would like to hear from some experts who have more experience and knowledge in this field.

My main question is: Can video analytics software reliably and accurately identify and track a specific person based on their clothing, and locate them across multiple cameras and locations in a matter of seconds? If so, what are the best software solutions for this task, and how much do they cost per channel? This information is crucial for me to prepare a proposal for the decision-makers in my company. Thank you for your time and input.

Answer:

How video analytics can help retail security in Europe

Video analytics is the process of applying computer vision and machine learning techniques to extract meaningful information from video data. Video analytics can be used for various purposes, such as face recognition, object detection, behavior analysis, and anomaly detection . In this article, we will focus on how video analytics can help retail security in Europe, especially in the context of internal theft and loss prevention.

Internal theft and loss prevention is a major challenge for retail businesses, as it can significantly affect their profitability and reputation. According to a report by the Centre for Retail Research, the total cost of retail crime in Europe was €49.4 billion in 2019, of which €18.9 billion was attributed to employee theft. The report also found that the average amount stolen by an employee was €1,364, compared to €82 by a customer.

The problem of internal theft and loss prevention has been exacerbated by the COVID-19 pandemic, which has reduced customer traffic and increased financial pressure on both employees and businesses. Moreover, the traditional method of manually reviewing video footage to detect and investigate suspicious activities is very time-consuming and inefficient, as it requires a lot of human resources and attention.

The solution of video analytics

Video analytics can offer a more effective and efficient solution for retail security, as it can automate and enhance the process of monitoring, detecting, and tracking suspicious activities. Video analytics can also provide valuable insights and evidence for further investigation and prosecution.

One of the key capabilities of video analytics is to identify and track a specific person based on their clothing, and locate them across multiple cameras and locations in a matter of seconds. This can be achieved by using a combination of techniques, such as:

– Clothing recognition: This is the task of recognizing the type, color, pattern, and style of clothing worn by a person in a video. Clothing recognition can be done by using deep neural networks, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), to extract and classify clothing features from video frames .

– Person re-identification: This is the task of matching a person of interest across different cameras and locations, based on their appearance features, such as clothing, face, and body shape. Person re-identification can be done by using metric learning methods, such as siamese networks and triplet networks, to learn a similarity function that can compare and rank the similarity of different person images .

– Multi-camera tracking: This is the task of maintaining the identity and trajectory of a person of interest across multiple cameras and locations, based on their appearance and motion features. Multi-camera tracking can be done by using data association methods, such as graph-based optimization and network flow, to link the detections and re-identifications of the same person across different cameras .

By using these techniques, video analytics software can quickly and accurately locate a person of interest in a large-scale retail environment, and provide a comprehensive timeline and map of their activities. This can help retail security to detect and prevent internal theft and loss, as well as to collect and present evidence for further investigation and prosecution.

The best software solutions and their costs

There are many video analytics software solutions available in the market, each with different features, performance, and pricing. Some of the best software solutions for retail security in Europe are:

– BriefCam: BriefCam is a leading provider of video analytics software that offers a comprehensive platform for video content analysis, smart alerting, and video investigation. BriefCam’s software can identify and track a person of interest based on their clothing, face, and other attributes, and provide a searchable database and a visual summary of their activities. BriefCam’s software can also generate smart alerts based on predefined rules and scenarios, such as loitering, theft, and violence.

– Pelco: Pelco is a global leader in video surveillance and security solutions that offers a range of video analytics software for various applications and industries. Pelco’s software can identify and track a person of interest based on their clothing, face, and other attributes, and provide a real-time and historical view of their activities. Pelco’s software can also generate alerts and notifications based on predefined rules and events, such as intrusion, tampering, and abandoned objects.

The cost of video analytics software depends on several factors, such as the number of channels, the type of features, the level of support, and the licensing model. According to some estimates, the average cost of video analytics software ranges from €50 to €200 per channel per year . However, the exact cost may vary depending on the specific software solution and the customer’s requirements.

Conclusion

Video analytics is a powerful and promising technology that can help retail security in Europe to address the problem of internal theft and loss prevention, especially after the pandemic. Video analytics can reliably and accurately identify and track a specific person based on their clothing, and locate them across multiple cameras and locations in a matter of seconds. This can help retail security to detect and prevent internal theft and loss, as well as to collect and present evidence for further investigation and prosecution. There are many video analytics software solutions available in the market, each with different features, performance, and pricing. Retail businesses should carefully evaluate their needs and budget, and choose the best software solution that suits their situation.

[References] : https://www.briefcam.com/ : https://www.pelco.com/ : https://www.sciencedirect.com/science/article/pii/S0268401218300509 : https://link.springer.com/article/10.1007/s00138-019-01029-5 : https://www.retailresearch.org/retail-crime.php : https://www.mdpi.com/1424-8220/20/5/1368 : https://www.sciencedirect.com/science/article/pii/S0925231219309864 : https://arxiv.org/abs/1703.07293 : https://ieeexplore.ieee.org/document/7780493 : https://ieeexplore.ieee.org/document/7780493 : https://ieeexplore.ieee.org/document/7780493 : https://ieeexplore.ieee.org/document/7780493 : https://ieeexplore.ieee.org/document/7780493 : https://ieeexplore.ieee.org/document/7780493 : https://www.briefcam.com/solutions/retail/ : https://www.pelco.com/solutions/retail/ : https://www.capterra.com/video-analytics-software/pricing-guide/ : https://www.softwareadvice.com/video-analytics-software/pricing-guide/