People counter

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

A people counter is a device used to measure the number and direction of people traversing a certain passage or entrance per unit time. The resolution of the measurement is entirely dependent on the sophistication of the technology employed. The device is often used at the entrance of a building so that the total number of visitors can be recorded. Many different technologies are used in people counter devices, such as infrared beams, computer vision, thermal imaging and pressure-sensitive mats.[1]

Reasons for use[edit]

Retail[edit]

Retail Management Dasboard for people counter and conversion rate

There are various reasons for counting people. In retail stores, counting is done as a form of intelligence-gathering. The use of people counting systems in the retail environment is necessary to calculate the conversion rate, i.e., the percentage of a store's visitors that makes purchases. This is the key performance indicator of a store's performance and is superior to traditional methods, which only take into account sales data. Together, traffic counts and conversion rates how a store arrived at sales, e.g., if year-over-year sales are down, did fewer people visit the store, or did fewer people buy? Accurate visitor counting is also useful in the process of optimizing staff shifts; Staff requirements are often directly related to density of visitor traffic and services such as cleaning and maintenance are typically done when traffic is at its lowest. More advanced People Counting technology can also be used for queue management and customer tracking. Although traffic counting is widely accepted as essential for retailers, it is estimated that less than 25% of major retailers track traffic in their stores.[citation needed]

Shopping mall marketing professionals rely on visitor statistics to measure their marketing. Often, shopping mall owners measure marketing effectiveness with sales, and also use visitor statistics to scientifically measure marketing effectiveness. Marketing metrics such as CPM (Cost Per Thousand) and SSF (Shoppers per Square Foot) are performance indicators that shopping mall owners monitor to determine rent according to the total number of visitors to the mall or according to the number of visitors to each individual store in the mall.

Occupancy[edit]

For safety, public locations are often rated to hold a certain number of people. Accurate people counting is used to ensure that the building is below the safe level of occupancy. Although, no people counting system is 100% accurate and therefore must not be entirely relied upon for the purposes of health & safety, an electronic people counting system offers a relatively accurate means of managing capacity.

Non-profit organizations[edit]

Many non-profit organizations use visitor counts as evidence when making applications for finance. In cases where tickets are not sold, such as in museums and libraries, counting is either automated, or staff keep a log of how many clients use different services.

Technologies[edit]

Modern people counting systems use many different technologies, each with its own advantages and disadvantages. The main types are listed below.

Tally counter[edit]

Main article: Tally counter

A hand-held tally-counter, sometimes called a clicker-counter, would be used; one press per person. To reset the counter, one would have to turn a knob, resetting most counters' display to "0000".

Infrared beams[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 or can also be linked to a PC or send data via wireless links and GPRS. 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. Dual beam units are also available from some suppliers and can provide low cost directional flow 'in' and 'out' data. Accuracy depends highly on the width of the entrance monitored and the volume of traffic.

Horizontal Beam Counters usually require a receiver or a reflector mounted opposite the unit with a typical range up to 6 metres (20 ft), although range finding beam counters which do not require a reflector or receiver usually have a shorter range of around 2.5 metres (8 ft 2 in).

Vertical beams are somewhat more accurate than horizontal, with accuracies of over 90% possible if the beams are very carefully placed. Typically they do not give 'in and out' information, although some directional beams do exist.

Advantages:

  • Inexpensive
  • Simple to fit

Disadvantages:

  • Most basic beam sensors are limited to non-directional counts
  • Can't discern people walking side-by-side
  • Cannot count high volume, uninterrupted traffic
  • High potential to become blocked by people standing in an entrance or by merchandise or displays
  • Infra-red beam counters may be negatively affected when subject to direct sunlight

Computer vision[edit]

Computer vision systems typically use either a closed-circuit television camera or IP camera to feed a signal into a computer or embedded device. Some computer vision systems have been embedded directly into standard IP network cameras. This allows for distributed, cost efficient and highly scalable systems where all image processing is done on the camera using the standard built in CPU. This also dramatically reduces bandwidth requirements as only the counting data has to be sent over the Ethernet.

Accuracy varies between systems and installations as background information needs to be digitally removed from the scene in order to recognize, track and count people. This means that CCTV based counters can be vulnerable to light level changes and shadows, which can lead to inaccurate counting. Lately, robust and adaptive algorithms has been developed that can compensate for this behavior and excellent counting accuracy can today be obtained for both outdoor and indoor counting using computer vision.[citation needed]

Advantages:

  • High accuracy, in correct conditions sometimes over 95%[citation needed]
  • Directional information
  • Flexible in customization
  • Highly scalable when embedded in IP cameras
  • Integration with other systems
  • Networkable to cover wide entrances
  • Possible to anonymize images to avoid people recognition.

Disadvantages:

  • Higher cost than beam systems
  • Lower lifetime and higher power consumption than thermal systems
  • Some systems require PCs are not fully embedded
  • Less simple implementation than beam systems
  • Accuracy can be affected by shadow, floor background, differing light levels

Thermal imaging[edit]

Thermal imaging systems use array sensors which detect heat sources, rather than using cameras as in computer vision systems. These systems are typically implemented using embedded technology and are mounted overhead for high accuracy. Since thermal imaging systems detect the heat emitted by people, they can be susceptible to external weather conditions that reduce the amount of heat emitted from a person walking in from an outdoor environment.

Advantages:

  • Directional information
  • Not affected by differing light levels
  • Can count in complete darkness
  • Non-intrusive, usually ceiling mounted
  • Identifiable images of people are not taken
  • High accuracy, in correct conditions over 98%[citation needed]
  • Very long lifetime - MTBF >25 years
  • Highly scalable, fully embedded IP Systems
  • Networkable to cover wide entrances

Disadvantages:

  • Higher cost than beam systems
  • Lower field of view than video systems
  • Cannot be used with ceiling heights below 2.2m
  • Susceptible to external weather conditions
  • Hard to know exactly how large an area or “spot” the sensor is “seeing” and measuring[2]
  • Takes considerable training to get the most from a thermal imager, which can be quite costly[3]

Artificial intelligence[edit]

This system employs multiple IR transceivers to create a count zone at ankle height. The artificial intelligence counters function in a similar way to the human brain, in other words, each event is evaluated in terms of features to determine the correct outcome i.e. count per direction. As a person passes the count zone a pattern is generated. The onboard processor extracts the features of the pattern and based on what it has been taught makes a decision regarding the event by brute force calculation.

AI using Pattern Recognition Technology (SRT)

Advantages:

  • Accuracy of 96% or higher[citation needed]
  • Directional information
  • Discriminates between human and non-human objects
  • Sensors can count in outdoor environments
  • Can count in all lighting conditions
  • Can count in complete darkness

Disadvantages:

  • Larger, more obtrusive design than other types of sensing technology.
  • High potential to become blocked by people standing in an entrance or by merchandise or displays.
  • Cannot count high volume, uninterrupted traffic

3D camera technology[edit]

Various technologies exist for acquiring a 3D image of the scene:

  • 2 video cameras in order to reproduce the human 3D vision (also known as stereo-vision)

Even under difficult conditions it achieves a counting accuracy of 98%[citation needed]. Unfavorable, variable lighting or the casting of shadows does not affect the counting accuracy. The quality is based on image processing and 3D camera technology.

The active technologies (those emitting light) have some difficulties in outdoor environments with direct or indirect sunlight, but the raw 3D point cloud is generally more robust for further processing such as counting applications.

The areas of application for such people counters are varied. They can be used in the public transport sector to record the capacity utilization and therefore to improve efficiency, or in stores and at trade fairs to record the flow of visitors. They also make the anonymous collection and analysis of visitor and object movements possible and help to optimize staff planning and to improve energy and environment management. The embedded design of the device combines the detection, evaluation and passing on of data in a compact housing.

References[edit]