People Counting

## Overview

People Counting is a dataset for object detection tasks - it contains People annotations for 1,198 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).

Crowd Counting Dataset

The dataset includes images featuring crowds of people ranging from 0 to 5000 individuals. The dataset includes a diverse range of scenes and scenarios, capturing crowds in various settings. Each image in the dataset is accompanied by a corresponding JSON file containing detailed labeling information for each person in the crowd for crowd count and classification.

Types of crowds in the dataset: 0-1000, 1000-2000, 2000-3000, 3000-4000 and 4000-5000 This… See the full description on the dataset page: https://huggingface.co/datasets/TrainingDataPro/crowd-counting-dataset.

People Counting System Market Report is Segmented by Offering (Hardware, Software, Services), Sensor Technology (Infrared Beam, Thermal Imaging, Video-Based and More), Deployment Mode (On-Premise, Cloud), Connectivity (Wired, Wireless, LP-WAN), End-User Vertical (Retail, Malls, Transportation, Hospitality and More), and by Geography. The Market Forecasts are Provided in Terms of Value (USD).

People Counting Yolov8

## Overview

People Counting Yolov8 is a dataset for object detection tasks - it contains Person annotations for 223 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Crowd counting from an image is a highly challenging task due to occlusion, low quality, and scale variation of objects. With the development of deep learning techniques, various crowd counting methods have been proposed in response to this challenge. This model uses state-of-the-art method to solve the crowd counting problem.Using the modelFollow the guide to use the model. Before using this model, ensure that the supported deep learning libraries are installed. For more details, check Deep Learning Libraries Installer for ArcGIS.Fine-tuning the modelThis model cannot be fine-tuned using ArcGIS tools.Input8 bit, 3-band (RGB) oriented imagery (preferably JPEG, JPG format with resolution less than 2000x2000 pixels).OutputFeature class with the number of classes as count of people.Applicable geographiesThis model is expected to work well in all regions globally. However, results can vary for imagery that are statistically dissimilar to training data.Model architectureThis model is based on the DM-Count model which uses the Distribution Matching for Crowd Counting architecture by Boyu Wang, Huidong Liu, Dimitris Samaras and Minh Hoai.Accuracy metricsThe average PSNR and SSIM over the QNRF test set are 40.65 and 0.55 respectively.Training dataThe model has been trained on the UCF-QNRF dataset.Sample resultsHere are a few results from the model.CitationsH. Idrees, M. Tayyab, K. Athrey, D. Zhang, S. Al-Maddeed, N. Rajpoot, M. Shah, Composition Loss for Counting, Density Map Estimation and Localization in Dense Crowds, in Proceedings of IEEE European Conference on Computer Vision (ECCV 2018), Munich, Germany, September 8-14, 2018.

The global people counting system market was valued at USD 912.3 million in 2021 and is expected to grow at a CAGR of 9.5% during the forecast period.

Here are a few use cases for this project:

1. Retail Analytics: Store owners can use the model to track the number of customers visiting their stores during different times of the day or seasons, which can help in workforce and resource allocation.

2. Crowd Management: Event organizers or public authorities can utilize the model to monitor crowd sizes at concerts, festivals, public gatherings or protests, aiding in security and emergency planning.

3. Smart Transportation: The model can be integrated into public transit systems to count the number of passengers in buses or trains, providing real-time occupancy information and assisting in transportation planning.

4. Health and Safety Compliance: During times of pandemics or emergencies, the model can be used to count the number of people in a location, ensuring compliance with restrictions on gathering sizes.

5. Building Security: The model can be adopted in security systems to track how many people enter and leave a building or a particular area, providing useful data for access control.

Dataset

This dataset was created by arpita

Contents

People Count

## Overview

People Count is a dataset for object detection tasks - it contains YOLOv8 annotations for 3,564 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

The retail people counting market, valued at $1556 million in 2025, is experiencing robust growth, projected to expand at a compound annual growth rate (CAGR) of 8.7% from 2025 to 2033. This expansion is fueled by several key factors. Firstly, the increasing adoption of advanced technologies like AI-powered video analytics, WiFi and Bluetooth sensing, and infrared sensors provides retailers with more accurate and granular data on customer traffic patterns. This allows for optimized store layouts, staffing levels, and marketing campaigns, leading to improved operational efficiency and enhanced customer experiences. Secondly, the growing demand for data-driven decision-making across the retail sector is driving the adoption of people counting systems. Retailers are increasingly realizing the importance of understanding customer behavior to personalize their offerings and optimize their strategies for better profitability. Finally, the increasing availability of affordable and user-friendly people counting solutions, including cloud-based platforms and mobile applications, is making this technology accessible to a wider range of businesses, from small and medium-sized enterprises (SMEs) to large multinational corporations. While the market faces challenges such as the initial investment costs associated with implementing these systems and concerns about data privacy, these are being mitigated by the long-term return on investment (ROI) generated through optimized operations and improved sales conversions. The market is segmented by application (SMEs and large enterprises) and technology (Wi-Fi and Bluetooth sensing, video-based counting, infrared sensors, time-of-flight sensors, and others). Key players in the market, including V-Count, Visionarea, Beonic (Blix), Retail Next, and ShopperTrak, are constantly innovating and expanding their product offerings to cater to the evolving needs of retailers. The competitive landscape is dynamic, with ongoing mergers, acquisitions, and the development of new technologies driving market evolution. The continued focus on enhancing the customer experience and leveraging data analytics will ensure sustained growth in the retail people counting market throughout the forecast period.

Crowd Density Dataset - Different Crowd Sizes

The dataset consists of 647 images of crowds, containing up to 11,000 individuals, annotated with keypoints for precise crowd counting and density estimation. It is designed for crowd counting tasks, particularly in crowded scenes settings, accommodating various sizes and challenges in estimating density. The dataset includes examples of both denser crowds and sparser crowds, enhancing counting accuracy for real-world applications in… See the full description on the dataset page: https://huggingface.co/datasets/UniDataPro/crowd-counting.

This dataset describes the number of people counted by infrared counting devices in the City of Ballarat. The information is collected in 15 minutes intervals. The intended use of the information is to inform the public of the number of people counted by infrared counting devices in the City of Ballarat. This dataset is updated automatically.

SAIVT-BuildingMonitoring

Overview

The SAIVT-BuildingMonitoring database contains footage from 12 cameras capturing a single work day at a busy university campus building. A portion of the database has been annotated for crowd counting and pedestrian throughput estimation, and is freely available for download. Contact Dr Simon Denman for more information.

Licensing

The SAIVT-BuildingMonitoring database is © 2015 QUT, and is licensed under the Creative Commons Attribution-ShareAlike 4.0 License.

Attribution

To attribute this database, use the citation provided on our publication at eprints:

S. Denman, C. Fookes, D. Ryan, & S. Sridharan (2015) Large scale monitoring of crowds and building utilisation: A new database and distributed approach. In 12th IEEE International Conference on Advanced Video and Signal Based Surveillance, 25-28 August 2015, Karlsruhe, Germany.

Acknowledgement in publications

In addition to citing our paper, we kindly request that the following text be included in an acknowledgements section at the end of your publications:

'We would like to thank the SAIVT Research Labs at Queensland University of Technology (QUT) for freely supplying us with the SAIVT-BuildingMonitoring database for our research'.

Installing the SAIVT-BuildingMonitoring Database

Download, join, and unzip the following archives

Annotated Data

  Part 1 (2GB, md5sum: 50e63a6ee394751fad75dc43017710e8)
  Part 2 (2GB, md5sum: 49859f0046f0b15d4cf0cfafceb9e88f)
  Part 3 (2GB, md5sum: b3c7386204930bc9d8545c1f4eb0c972)
  Part 4 (2GB, md5sum: 4606fc090f6020b771f74d565fc73f6d)
  Part 5 (632 MB, md5sum: 116aade568ccfeaefcdd07b5110b815a)


Full Sequences

  Part 1 (2 GB, md5sum: 068ed015e057afb98b404dd95dc8fbb3)
  Part 2 (2GB, md5sum: 763f46fc1251a2301cb63b697c881db2)
  Part 3 (2GB, md5sum: 75e7090c6035b0962e2b05a3a8e4c59e)
  Part 4 (2GB, md5sum: 34481b1e81e06310238d9ed3a57b25af)
  Part 5 (2GB, md5sum: 9ef895c2def141d712a557a6a72d3bcc)
  Part 6 (2GB, md5sum: 2a76e6b199dccae0113a8fd509bf8a04)
  Part 7 (2GB, md5sum: 77c659ab6002767cc13794aa1279f2dd)
  Part 8 (2GB, md5sum: 703f54f297b4c93e53c662c83e42372c)
  Part 9 (2GB, md5sum: 65ebdab38367cf22b057a8667b76068d)
  Part 10 (2GB, md5sum: bb5f6527f65760717cd819b826674d83)
  Part 11 (2GB, md5sum: 01a562f7bd659fb9b81362c44838bfb1)
  Part 12 (2GB, md5sum: 5e4a0d4bb99cde17158c1f346bbbdad8)
  Part 13 (2GB, md5sum: 9c454d9381a1c8a4e8dc68cfaeaf4622)
  Part 14 (2GB, md5sum: 8ff2b03b22d0c9ca528544193599dc18)
  Part 15 (2GB, md5sum: 86efac1962e2bef3afd3867f8dda1437)

To rejoin the invidual parts, use:

cat SAIVT-BuildingMonitoring-AnnotatedData.tar.gz.* > SAIVT-BuildingMonitoring-AnnotatedData.tar.gz

cat SAIVT-BuildingMonitoring-FullSequences.tar.gz.* > SAIVT-BuildingMonitoring-FullSequences.tar.gz

At this point, you should have the following data structure and the SAIVT-BuildingMonitoring database is installed:

SAIVT-BuildingMonitoring +-- AnnotatedData +-- P_Lev_4_Entry_Way_ip_107 +-- Frames +-- Entry_ip107_00000.png +-- Entry_ip107_00001.png +-- ... +-- GroundTruth.xml +-- P_Lev_4_Entry_Way_ip_107-20140730-090000.avi +-- perspectivemap.xml +-- ROI.xml

+-- P_Lev_4_external_419_ip_52 +-- ...

+-- P_Lev_4_External_Lift_foyer_ip_70 +-- Frames +-- Entry_ip107_00000.png +-- Entry_ip107_00001.png +-- ... +-- GroundTruth.xml +-- P_Lev_4_External_Lift_foyer_ip_70-20140730-090000.avi +-- perspectivemap.xml +-- ROI.xml +-- VG-GroundTruth.xml +-- VG-ROI.xml

+-- ...

+-- Calibration +-- Lev4Entry_ip107.xml +-- Lev4Ext_ip51.xml +-- ...

+-- FullSequences +-- P_Lev_4_Entry_Way_ip_107-20140730-090000.avi +-- P_Lev_4_external_419_ip_52-20140730-090000.avi +-- ...

+-- MotionSegmentation +-- Lev4Entry_ip107.avi +-- Lev4Entry_ip107-Full.avi +-- Lev4Ext_ip51.avi +-- Lev4Ext_ip51-Full.avi +-- ...

+-- Denman 2015 - Large scale monitoring of crowds and building utilisation.pdf +-- LICENSE.txt +-- README.txt

Data is organised into two sections, AnnotatedData and FullSequences. Additional data that may be of use is provided in Calibration and MotionSegmentation.

AnnotatedData contains the two hour sections that have been annotated (from 11am to 1pm), alongside the ground truth and any other data generated during the annotation process. Each camera has a directory, the contents of which depends on what the camera has been annotated for.

All cameras will have:

a video file, such as P_Lev_4_Entry_Way_ip_107-20140730-090000.avi, which is the 2 hour video from 11am to 1pm
a Frames directory, that has 120 frames taken at minute intervals from the sequence. There are the frames that have been annotated for crowd counting. Even if the camera has not been annotated for crowd counting (i.e. P_Lev_4_Main_Entry_ip_54), this directory is included.

The following files exist for crowd counting cameras:

GroundTruth.xml, which contains the ground truth in the following format: 

The file contains a list of annotated frames, and the location of the approximate centre of mass of any people within the frame. The interval-scale attribute indicates the distance between the annotated frames in the original video.

perspectivemap.xml, a file that defines the perspective map used to correct for perspective distortion. Parameters for a bilinear perspective map are included along with the original annotations that were used to generate the map.
ROI.xml, which defines the region of interest as follows:

This defines a polygon within the image that is used for crowd counting. Only people within this region are annotated.

For cameras that have been annotated with a virtual gate, the following additional files are present:

VG-GroundTruth.xml, which contains ground truth in the following format: 

The ROI is repeated within the ground truth, and a direction of interest (the tag) is also included, which indicates the primary direction for the gait (i.e. the direction that denotes a positive count. Each pedestrian crossing is represented by a tag, which contains the approximate frame the crossing occurred in (when the centre of mass was at the centre of the gait region), the x and y location of the centre of mass of the person during the crossing, and the direction (0 being the primary direction, 1 being the secondary). VG-ROI.xml, which contains the region of interest for the virtual gate

The Calibration directory contains camera calibration for the cameras (with the exception of ip107, which has an uneven ground plane and is thus difficult to calibrate). All calibration is done using Tsai's method.

FullSequences contains the full sequences (9am - 5pm) for each of the cameras.

MotionSegmentation contains motion segmentation videos for all clips. Segmentation videos for both the full sequences and the 2 hour annotated segments are provided. Motion segmentation is done using the ViBE algorithm. Motion videos for the entire sequence have Full in the file name before the extension (i.e. Lev4Entry_ip107-Full.avi).

Further information on the SAIVT-BuildingMonitoring database in our paper: S. Denman, C. Fookes, D. Ryan, & S. Sridharan (2015) Large scale monitoring of crowds and building utilisation: A new database and distributed approach. In 12th IEEE International Conference on Advanced Video and Signal Based Surveillance, 25-28 August 2015, Karlsruhe, Germany.

This paper is also available alongside this document in the file: 'Denman 2015 - Large scale monitoring of crowds and building utilisation.pdf'.

The nCounter sensors count the number of Wi-Fi enabled mobile devices in a specified zone. An nCounter can report the number of mobile devices that enter a zone, that leave a zone and the dwell time (in minutes). This provides insightful data on number of people in an area.It provides an indication of trends, activities and events within access point areas. It can be visualised on our People & Devices Counting dashboard.Anyone using this data does so at their own risk. To the full extent permitted by law, Council is released from and will in no way be liable to you or anyone else for any loss however caused (including through negligence) suffered directly or indirectly as a result of any reliance on this data. All intellectual property rights are reserved.

The AI people counter market is experiencing robust growth, driven by the increasing need for accurate and real-time foot traffic analytics across diverse sectors. The market's expansion is fueled by several key factors. Firstly, the rising adoption of AI-powered solutions in retail, office buildings, and public transportation offers businesses valuable insights into customer behavior, optimizing operational efficiency and enhancing the overall customer experience. Secondly, advancements in computer vision and infrared technology are leading to more sophisticated and reliable people counting systems, improving accuracy and reducing errors associated with traditional methods. Finally, the increasing affordability and accessibility of AI-based solutions are making them attractive to a wider range of businesses, regardless of size. We estimate the 2025 market size to be approximately $800 million, considering the prevalent growth in related technologies and the expanding adoption across various sectors. A conservative Compound Annual Growth Rate (CAGR) of 15% over the forecast period (2025-2033) is projected, leading to significant market expansion by 2033. Despite the positive outlook, certain restraints could impede market growth. Data privacy concerns surrounding the collection and usage of foot traffic data remain a challenge. The high initial investment costs associated with implementing AI people counting systems might deter smaller businesses. Furthermore, the accuracy of AI-based systems can be impacted by factors such as lighting conditions, obstructions, and crowd density, demanding continuous improvements in technology. Market segmentation reveals a strong demand for computer vision-based systems, due to their versatility and advanced analytical capabilities. The retail and office building sectors are leading adopters, followed by public transportation and other applications. Key players like V-Count, Hikvision, and others are driving innovation and competition in this dynamic landscape. Geographical expansion is also observed, with North America and Europe currently dominating the market share, while Asia Pacific is projected to experience significant growth due to rapid technological adoption and increasing urbanization.

This dataset brings together device, people, vehicle and bicycle counting data from multiple councils (Liverpool, Wollondilly, Penrith and Blue Mountains) and collected via different methods (nCounter sensors, CCTV cameras equipped with video analytics software and Telstra Purple WiFi infrastructure). The nCounter sensors and Purple WiFi count the number of Wi-Fi signals emitted by non-identifiable mobile devices within a specified proximity and performing certain filtering and processing. The data can also be visualised on the Pedestrians and Vehicles counting dashboard.Please note this data is indicative as sensors may from time to time provide incorrect data due to wear and tear or unforeseen circumstances. Anyone using this data does so at their own risk. To the full extent permitted by law, Council is released from and will in no way be liable to you or anyone else for any loss however caused (including through negligence) suffered directly or indirectly as a result of any reliance on this data. All intellectual property rights are reserved.

The Shanghaitech dataset is a large-scale crowd counting dataset. It consists of 1198 annotated crowd images. The dataset is divided into two parts, Part-A containing 482 images and Part-B containing 716 images. Part-A is split into train and test subsets consisting of 300 and 182 images, respectively. Part-B is split into train and test subsets consisting of 400 and 316 images. Each person in a crowd image is annotated with one point close to the center of the head. In total, the dataset consists of 330,165 annotated people. Images from Part-A were collected from the Internet, while images from Part-B were collected on the busy streets of Shanghai.

The global AI people counter market is experiencing robust growth, driven by the increasing need for real-time customer analytics and improved operational efficiency across various sectors. The market, estimated at $500 million in 2025, is projected to witness a Compound Annual Growth Rate (CAGR) of 15% from 2025 to 2033, reaching approximately $1.5 billion by 2033. This expansion is fueled by several key factors. Firstly, the rising adoption of AI-powered solutions across retail, hospitality, and transportation is significantly boosting demand. Businesses are leveraging these counters to gain valuable insights into customer behavior, optimize store layouts, and improve resource allocation. Secondly, advancements in computer vision and deep learning technologies are leading to more accurate and reliable people counting, further enhancing the appeal of these systems. Finally, the increasing affordability of AI-powered solutions is making them accessible to a wider range of businesses, accelerating market penetration. However, certain challenges hinder market growth. Initial investment costs associated with installing and integrating AI people counters can be significant, particularly for smaller businesses. Concerns regarding data privacy and security also need to be addressed to build trust and encourage wider adoption. Despite these restraints, the market is poised for significant expansion, driven by continuous technological innovation and the growing demand for data-driven decision-making across various industries. The competitive landscape is dynamic, with established players like Hikvision and V-Count alongside emerging companies like Plugger AI and Dragonfruit AI vying for market share. Future growth will likely be shaped by the development of more sophisticated analytics capabilities, integration with other business intelligence platforms, and the increasing focus on providing secure and privacy-compliant solutions.

The Gridded Population of the World, Version 3 (GPWv3): Population Count Grid, Future Estimates consists of estimates of human population for the years 2005, 2010, and 2015 by 2.5 arc-minute grid cells and associated data sets dated circa 2000. A proportional allocation gridding algorithm, utilizing more than 300,000 national and sub-national administrative Units, is used to assign population values to grid cells. The population counts that the grids are derived from are extrapolated based on a combination of subnational growth rates from census dates and national growth rates from United Nations statistics. All of the grids have been adjusted to match United Nations national level population estimates. The population count grids contain estimates of the number of persons per grid cell. The grids are available in various GIS-compatible data formats and geographic extents (global, continent [Antarctica not included], and country levels). GPWv3 is produced by the Columbia University Center for International Earth Science Information Network (CIESIN) in collaboration with Centro Internacional de Agricultura Tropical (CIAT).

Indoor People Counter Market Outlook

The global indoor people counter market size was valued at approximately USD 1.2 billion in 2023 and is projected to reach USD 3.5 billion by 2032, growing at a compound annual growth rate (CAGR) of 12.5% during the forecast period. This remarkable growth can be attributed to the increasing need for accurate people counting solutions across various industries, driven by advancements in technology and a heightened focus on enhancing customer experiences and operational efficiency.

One of the key growth factors for the indoor people counter market is the rising demand for data analytics in retail settings. Retailers are increasingly leveraging people counting systems to gain insights into customer behavior, optimize store layouts, and enhance marketing strategies. With the integration of advanced technologies such as artificial intelligence and machine learning, these systems are becoming more accurate and sophisticated, further driving their adoption. Additionally, the pressing need for real-time data to make informed business decisions is propelling the market's growth.

Another significant factor contributing to market growth is the adoption of people counting solutions in transportation hubs, such as airports and train stations. Transportation authorities are utilizing these systems to manage passenger flow, ensure safety, and improve infrastructure planning. The growing emphasis on smart city initiatives and the expansion of public transport networks in urban areas are further fueling the demand for indoor people counters. Moreover, government regulations and security concerns are prompting the deployment of these systems in public spaces.

The hospitality industry is also a major contributor to the growth of this market. Hotels and event venues are increasingly employing people counting technologies to manage crowd control, enhance guest experiences, and optimize resource allocation. The ability to monitor occupancy levels in real time is crucial for maintaining safety standards and ensuring compliance with health regulations, particularly in the context of the COVID-19 pandemic. As the hospitality sector continues to recover and adapt to new norms, the adoption of indoor people counters is expected to rise.

Regionally, North America holds a significant share of the indoor people counter market, driven by the presence of major technology providers and extensive retail and transportation infrastructure. The region's focus on adopting cutting-edge technologies and the high penetration of advanced analytics solutions are key growth drivers. Meanwhile, the Asia Pacific region is witnessing rapid growth due to the increasing urbanization, expanding retail sector, and government initiatives promoting smart city projects. Europe also presents substantial growth opportunities, supported by the strong presence of retail giants and technological advancements.

In addition to the indoor people counter market, the use of Fish Counting Systems is gaining traction, particularly in the environmental and aquaculture sectors. These systems are designed to accurately monitor fish populations in rivers, lakes, and fish farms, providing valuable data for conservation efforts and sustainable fishery management. By utilizing advanced technologies such as sonar and video imaging, fish counting systems can offer precise insights into fish behavior, migration patterns, and population dynamics. This data is crucial for ensuring the health of aquatic ecosystems and supporting the sustainable management of fish resources. As environmental concerns and the demand for sustainable practices grow, the adoption of fish counting systems is expected to increase, offering new opportunities for technological advancements and market expansion.

Component Analysis

In the indoor people counter market, the component segment is divided into hardware, software, and services. Each of these components plays a crucial role in the overall functionality and effectiveness of people counting systems. Hardware components include sensors, cameras, and other physical devices used to detect and count individuals. This hardware is often integrated with sophisticated software that processes the data collected, providing valuable insights and analytics. Services encompass installation, maintenance, and technical support, ensuring the smooth operation of the systems.

The hardware segment is anticipated to hold the