People counter

From Wikipedia, the free encyclopedia
Jump to: navigation, search

A people counter is an electronic device that is used to measure the number of people traversing a certain passage or entrance.[1] Currently, the people counter industry has been forecasted by ABI Research[2] to grow upward of 3 billion USD.

Use cases[edit]

Retail stores[edit]

Conversion rate: People counting systems in the retail environment are used to calculate the conversion rate, which is the percentage of total visitors versus the number that make purchases.

Marketing effectiveness: Shopping mall marketing professionals rely on visitor statistics to measure the effectiveness of the current marketing campaign. Often, shopping mall owners measure marketing effectiveness with the same conversion rate as retail stores.[3]

Staff planning: Retailers can use the different business metrics in order to determine their staffing allocation. Accurate visitor counting is also useful for optimizing staff shifts. Staff requirements are often directly related to the density of visitor traffic, and services such as cleaning and maintenance are typically undertaken when traffic is at its lowest.[4]

Shopping malls[edit]

Monitoring of high-traffic areas: Shopping centers use people counters to measure the number of visitors in a given area. People counters also assist in measuring the areas where people tend to congregate. The areas where people tend to gather are often charged higher rent.[5]

Determining popularity of particular brands: Shopping malls are prevalent in terms of leasing their units and store lot to only the most popular brand. This is due to the predominant stream of revenue a brand will be able to generate if there is enough demand for it. People counters help shopping malls determine footfall pattern and traffic. With people counters, shopping mall owners are able to determine the flow of traffic per each customer, and which areas and entrances are widely used throughout the whole mall.[6]

Business metrics[edit]

People counters are mostly used to measure different business metrics. While there are many different types of people counters and each model varies in the metrics that is offered, people counters will include some or all of the following metrics to an extent.

Sample footfall report that reports visitor count in a particular day

Footfall[edit]

Footfall measures the number of people who enter a shop or business in a particular period of time.[7]

Conversion rate[edit]

The window conversion rate is the percentage of shoppers who enter a store over in relation to the number of people who walk by it. With WiFi counting, shops can estimate the number of people who walk past a store. A more accurate method is video counting. While revenue and footfall are important, the number of people who walk past a store often reflects the potential of the store location. The window conversion rate often depends on the attractiveness of the shop window design and the effectiveness of marketing campaigns.[8]

Visit duration[edit]

Visit duration is the amount of time visitors stay in a venue. WiFi counting has the ability to track both the time a person carrying a smartphone has entered the venue, as well as when that same person has left the venue.[9]


Zone counting/traffic flow[edit]

This metric, similar to the bubble map and heat map, allows the user to see the flow of traffic in a shopping mall and to analyze the engagement percentile across the entire compound. The zone counting and traffic flow will be useful for mall owners to determine which section of the mall is more popular, and can choose to maintain the sector accordingly. With the traffic flow diagram, the mall owner will be able to determine what is the most popular district of the mall, and may choose to lease their rental areas based on demand.[10]

Bubble map/heat map[edit]

This metric tracks the user engagement per districts, sections, and courts of a compound. The bubble map/heat map allows users to analyze the number of engagement in percentage across the entire compound at a period of time. The bubble map and heat map functions similarly, the only difference is the methodology of display. A heat map shows the engagement level through the use of colors, with warmer colors showing more engagement, whereas a bubble map is more straight forward. The bubble map shows engagement in percentile and circumference of the bubble drawn.

Outside traffic[edit]

Outside traffic allows retailers to determine the number of people passing by the retail store on any given day. With the measurement of outside traffic, retailers are able to estimate how many potential customers one location will be able to bring to the business.[11] Outside traffic is able to help businesses to determine the ideal business location to set up their business, while also being able to determine the worth and value of the property due to the number of visitors that is found around the premise.

Returning customers[edit]

This metric looks at the number of people entering a store who had visited the store previously. WiFi counting has the ability to remember the unique WiFi signal ID emitted by shoppers, allowing the system to detect if a shopper has previously visited the store. This is due to the fact that WiFi analytics function by smartphones emits a unique identification signal that allows people counters to easily distinguish whether the signal has been detected once before.[12]

Current technology[edit]

Many different technologies are used in people counting devices, such as infrared beams, thermal imaging, computer vision, and WiFi counting.[13] Recently, pressure sensitive sensors that count walk-ins based on the number of footsteps on a pressure sensitive platform or mat have been used as well.

4th generation: Video and wifi analytics with video footage (2017 to present)[edit]

The 4th generation of people counters combines the feature of both video or thermal counting and WiFi counting from the 3rd generation of people counter. Additionally, all 4th-generation people counters have a built-in recorder that allows the user to review the accuracy of the people counter.

Video counting[edit]

People counter seamlessly installed in a retail store

Computer vision works via an embedded device, reducing the network bandwidth usage, as only the number of people must be sent over the network. Adaptive algorithms have been developed to provide accurate counting for both outdoor and indoor counting using video counting. Multi-layer background subtraction, based on color and texture, is considered the most robust algorithm[14] available for varying shadows and lighting conditions.[15] With the advances in image processing, video counting can achieve 98% accuracy in some lighting environments.[16] The use of artificial intelligence and pattern recognition functions is expected to further enhance its accuracy.[17]

WiFi counting[edit]

WiFi counting uses a WiFi receiver to pick up unique WiFi management frames emitted from smartphones within range.[18] While not all people carry a smartphone, WiFi counting can produce statistically significant metrics with a large enough sample size. Modern mobile operating systems, such as Apple's iOS9 and Android 6.0 Marshmallow, use MAC rotation schemes which makes WiFi counting more challenging without some sophisticated algorithms.[19][20]

Video verification[edit]

The 4th generation of people counters includes an option for users to review the authenticity and integrity of the data provided by their people counter. The user will be able to verify the accuracy of the counter and make informed business decisions accordingly, factoring in all the disparity of data. They are now used in the 2nd generation.

Seamless integration with store environment[edit]

People counters seamlessly integrate with the store environment in order to minimize obstruction and disruption of the store environment. Furthermore, since people counters may be easily mistaken for a surveillance camera, shoppers may feel uneasy and distracted during the experience. Causing a reverse situation where the purpose of people counters is to aid in the analysis of shopping behaviour, as opposed to discouraging it.

History[edit]

Prior to electronic people counters[edit]

The most traditional people counter device used before the inception of electronic people counters were manual people counters. These people counters required a store employee to stand near the entrance of the store and click on a pedometer device to track each person that enters the store. This form of manual people counter was considered to be inaccurate due to the multiple chances of a person making human errors, as well as an inefficient usage of human resource. Pressure sensitive sensors that count walk-ins based on the number of footsteps on a pressure sensitive platform or mat have been used as well.

The simplest form of counter in which a single, horizontal infrared beam across an entrance counts when a person or object passes and breaks its beam

1st generation: Infrared beam counters (2002 to 2004)[edit]

The simplest form of counter is a single, horizontal infrared beam across an entrance which is typically linked to a small LCD display unit at the side of the doorway. Such a beam counts a 'tick' when the beam is broken, therefore it is normal to divide the 'ticks' by two to get visitor numbers. Beam counters usually require a receiver or a reflector mounted opposite the unit with a typical range from 2.5 metres (8 ft 2 in) to 6 metres (20 ft). Despite the limitations, infrared counters are still widely used due to the low cost and simplicity of installation.[21]

Thermal imaging systems use array sensors which detect heat sources from human body.

2nd generation: Thermal counters (2005 to 2011)[edit]

Thermal imaging systems use array sensors that detect heat sources. These systems are typically implemented using embedded technology and are mounted overhead for high accuracy.

Before the advance of computer technology that allows complex algorithms to perform video counting, thermal counters were the main choice for most businesses.[citation needed] They are fairly accurate; however, they do have limitations, such as:

  1. Thermal counters cannot be mounted on a high ceiling without using a narrow-angle lens
  2. Thermal counters have difficulty measuring the dwell time of targets beyond a few seconds
  3. It is difficult to verify the accuracy of the counter
  4. Accuracy is reduced in places with significant variations in thermal conditions

Thermal people counters are still a popular choice thanks to the strengths of the technology:

  1. Without an optical sensor they can be used in places where legal/privacy concerns prohibit the recording of images of people
  2. Thermal sensors are much less affected by low/variable light conditions than optical sensors, and can even be used in complete darkness
  3. Patterns on flooring that may disrupt optical sensors do not affect thermal sensors

3rd generation: Video and WiFi counting (2012 to 2016)[edit]

There are two types of 3rd-generation people counters. Video counters use complex algorithms and camera imaging to count the number of people directly from a video tape. Wifi counting functionality collects WiFi probe request signals from shoppers' smartphones, allowing data to be collected on those not in the store. This adds a number of important metrics for businesses, especially for the retail industry, such as the ability to determine how effective a window marketing campaign is.

Shortcomings[edit]

People counters are all-important tools for bricks and mortar to calculate their sales conversion and revenue. Additionally, people counters allow retailers to determine the most efficient timing for staffing options. However, a lot of retail stores have yet to adopt people counters in their premises due to their various shortcomings.

Data inaccuracy[edit]

The inaccuracy of the people counter is a major factor that prevents most retail stores from adopting one in their premise. Without any visibility of how the algorithm of people counting works, it is difficult to quantify the accuracy of the people counter.[22] For example, if three people walk together as a group into the building, are they counted as three people separately or are they considered as one person? This issue is not as prevalent with recent models of people counters as there is now a video recorder built into each people counter that allows the user to effectively determine the accuracy of the people counters.

Children counting[edit]

Many people counters are unable to distinguish between children and adults. This creates an inconsistency in the sales data once the user has imported their sales data from their ePoS system. Many businesses that use people counters do not like the counting of children due to their lack of purchasing power. Without any stream of income, the counting of children will alter the value of the sales conversion rate and make it seem lower than it really is.

Staff counting[edit]

People counters are usually conducted via recorded video footage that is able to analyse the field of depth or changes in the background to determine the movement of visitors in an environment. There is currently no people counter to date that is able to successfully distinguish the staff from a store from visitors since people counting technology is not advanced enough to have facial recognition. Without a method to exclude the counting of staff from a particular entrance, many users of people counters fear that the generated data value will be inaccurate and does not accurately reflect the performance of the retail store.

See also[edit]

References[edit]

  1. ^ Name, No. "Arduino-DIY Laser / IR Person Counter". Instrutables. Retrieved 21 August 2017. 
  2. ^ "People Counting Retail Market Undergoing $3 Billion Technology Evolution". ABI Research. Retrieved 18 August 2017. 
  3. ^ Editors, Allbusiness. "Metrics for measuring ad campaign effectiveness". AllBusiness. Retrieved 17 August 2017. 
  4. ^ "Using marketing insights to optimize staffing". MyCustomer. Retrieved 18 August 2017. 
  5. ^ Fathi, Nader. "How Shopper Analytics ensure mall thrives?". Chain Store Age. Retrieved 17 August 2017. 
  6. ^ D'Mello, Sandhya. "Malls in Dubai use sensors to count crowds". www.khaleejtimes.com. Retrieved 2017-08-18. 
  7. ^ Dictionary, Oxford. "footfall". Oxford Dictionary. Retrieved 17 August 2017. 
  8. ^ Dillon, Chris (30 July 2015). "Sunder Sandhe's tech game changer". 
  9. ^ "Wi-Fi Location Analytics" (PDF). Information Commissioners Office. Retrieved 18 August 2017. 
  10. ^ Lowry, James. "How to Use a Traffic Study to Select a Retail Site" (PDF). Women's Enterprise Centre. Retrieved 21 August 2017. 
  11. ^ Marsan, Jeremy. "How To Determine Foot Traffic & Use Data To Pick A Business Location". Fit Small Business. Retrieved 18 August 2017. 
  12. ^ Clifford, Stephanie; Hardy, Quentin. "Attention, Shoppers: Store Is Tracking Your Cell". NY Times. Retrieved 18 August 2017. 
  13. ^ Paul C. Box, Joseph C. Oppenlander (1976), Manual of traffic engineering studies, Institute of Transportation Engineers, p. 17, retrieved December 21, 2010 
  14. ^ Yao, Jian; Odobez, Jean-Marc. "Multi-Layer Background Subtraction Based on Color and Texture" (PDF). Semantics Scholar. Retrieved 21 August 2017. 
  15. ^ Jian Yao; Jean-Marc Odobez (2007). Multi-Layer Background Subtraction Based on Color and Texture (PDF). The CVPR Visual Surveillance Workshop (CVPR-VS). Minneapolis, MN, US: IEEE. doi:10.1109/CVPR.2007.383497. 
  16. ^ "How a CCTV People Counting System Works". Retail Sensing. Retrieved 15 April 2016. 
  17. ^ Hsu, Jeremy. "Computer Count of huge crowd now possible". IEEE Spectrum. Retrieved 18 August 2017. 
  18. ^ Wenz, John. "How Wi-Fi Can Count the People in a Room Without Tracking Their Phones". Popular Mechanics. Retrieved 17 August 2017. 
  19. ^ "Security and Privacy Changes in iOS 9". 2015. 
  20. ^ Amadeo, Ron. "Android 6.0 Marshmallow, thoroughly reviewed". arstechnica.com. 
  21. ^ Kajala, L.; Almik, A.; Dahl, R.; Diksaito, L.; Erkkonen, J.; Fredman, P.; Jensen, F.; Sondergaard, K,; Sievaner, T. (2007). Visitor Monitoring in Nature Area - a manual based on experiences from the Nordic and Baltic Countries. Sweden: TemaNord. p. 46. Retrieved 21 August 2017. 
  22. ^ Clements, Bob. "Five Reasons Why Retailers Don't Use Traffic to Drive Labor". Axium. Retrieved 18 August 2017.