People Analytics
How it Works
People Counting
Nodeflux People Counting is a Vision Artificial Intelligence Analytics Software that allows you to know the number of people in a particular area with multiple regions of interest. This can be done to capture people traffic, like at the front door, gate, elevator, lift, lobby, etc. Using Nodeflux People Counting, organizations can seamlessly gather the number of visitor data and analyze peak hours in each respective area.
To verify that your camera can work with our People Counting analytic, try performing a test. Make sure that our analytic can detect people (you can see a bounding box around that person) and the bounding box can follow the person till crossing the counter line.
Our model requires to detect person with height above 200 pixels on frame or the a quarter from height your display if the frame is 1080p. Less than that, the result might varies depends on the how the person visually presented.
People Dwelling
Nodeflux People Dwelling is a Vision Artificial Intelligence Analytics Software that allows you to know the number of people in a particular area and how long they stay at the same position. This can be used to capture people's traffic, like at the front door, gate, elevator, lift, lobby, in front of shelves, and etc. Using Nodeflux People Dwelling, organizations can seamlessly gather the number of visitor data, how long they stay at the same position and analyze their behavior in each respective area.
People Density
Nodeflux People Density is a Vision AI analytics software which aims to measure the number of people at small to low medium density in an area of interest. The analytics work within the common practice is to count each person automatically to get an insightful report. The inference result will provide you the exact number of people in a particular area of interest at small to low medium density.
People Analytics with Attributes
Nodeflux People Analytics with Attributes is same as the People Analytics only explained above excluding the logic of People Density. The Vision AI analytics software provides the enrichment data of person attributes within the person detection in People Counting and People Dwelling.
The attributes data includes the broad range of enrichment data which are:
Age: Teenager, Adult, Elderly
Gender: Female, Male
Bags: HandBad, ShoulderBag, Backpack
Hat: Hat and No Hat
Glasses: Glasses and No Glasses
Upper Wear: UpperStride, UpperLogo, UpperPlaid, UpperSplice
Lower Wear: LowerStripe, LowerPattern, LongCoat, Trousers, Shorts, Skirt&Dress
Since Nodeflux People Analytics with Attributes utilizes the snapshot platform, an additional docker-compose.yml file needs to be deployed alongside the other below services. The following directive link outlines the installation process for deploying the docker-compose.yml file specific to the people attribute pipeline of snapshot platform:
Architecture
Visionaire People Analytics utilizes a combination of other services:
Postgres - For database,
Docker Snapshot - For People Analytics with Attributes related,
Visionaire Docker Stream - For video stream processing,
Visionaire Dashboard (Optional) - Built-in dashboard for visualization.
Camera Specification
Frame Resolution
1 - 4K
720p - 1080p
Codec
H.264
H.264
Codec Profile
Baseline
Baseline
Bit Rate Mode
Variable (VBR)
Variable (VBR)
Target Bitrate
2 - 6 Mbps
1 - 4 Mbps
FPS Rate
25 - 30
15 - 30
Bandwidth
4 Mbps
4 Mbps
Exposure Time/Shutter Speed
1/30 - 1/120s
1/30s - 1/120s
Camera Placement
1. Short Range View (Up to 5 meters)
Camera Placement: Ideal height is around 2 to 2.5 meters except for crowd estimation around 5 m - 10 m. Place cameras to capture areas like entrances, counters, or specific zones where individuals will be in close proximity.
Camera Type and Specs: Use wide-angle lenses for a broader field of view. A resolution of 1080p or higher ensures detailed images.
Lighting: Uniform, soft lighting to minimize shadows and glare. Artificial lighting should complement natural light.
Analytics Focus: Ideal for detailed behavioral analysis and individual interactions.
2. Medium Range View (5 - 15 meters)
Camera Placement: Height around 3 to 4 meters except for crowd estimation around 5 m - 10 m. Cameras should overlook areas like lobbies, hallways, or open office spaces.
Camera Type and Specs: Standard lenses are suitable. A resolution of 1080p to 4K is recommended to maintain detail at this range.
Lighting: Balanced lighting is crucial. Avoid areas with significant light variation (like directly under skylights).
Analytics Focus: Suited for tracking movement patterns, group interactions, and general behavior analysis.
3. Additional Considerations
Angle of Capture: Should be as direct as possible. Avoid steep angles to prevent distortion and loss of detail.
Data Privacy: Ensure compliance with privacy laws and regulations. Inform individuals about surveillance and data usage.
Environmental Factors: Outdoor cameras need to be weatherproof and capable of adjusting to changing light conditions.
Integration with Analytics Software: Ensure cameras are compatible with your analytics software for seamless data processing and analysis.
Sample View
View to Avoid
Please avoid a CCTV view when there are a lot of mannequins, banner/digital ads which shows a person or hanged clothes in monitored area to prevent false detection.
Deployment Schema
Last updated
