Features and Benefits
- Counting based on the direction of movement, enabling users to count number of entries and exits in an area of interest.
- High performance software to run real-time on VGA resolution capture at 30 frames per second
- Analytics completely embedded on edge device (i.e. BLIP) with no processing required to happen on cloud or other host device
- Accurate counting (~90%) in low to medium density scenarios.
- Image processing and Machine learning algorithms minimizes reflections or shadow problems
- Multiple connectivity options to send the output metadata on UART, SPI and WiFi. Open, documented protocol helps in seamless integration to other microcontroller or host devices on SPI/UART.
- The product supports cloud connectivity to ThingWorx cloud on WiFi. The metadata pushed to the Cloud from this product can enable services to provide powerful business analytic tools and reporting
- Easy to use GUI application for installation, visualization, evaluation and configuration.
- Target Processor - ADSP-BF7xx
- Release format - Object binary executable
- Input format - UYVY422 raw progressive or Y only
- Output –
- Number of Exits and Entries
- Features of counted and tracked people such as Area, Bounding Rectangle Box, Centroid
- Operational modes
- Configuration - Easy to use GUI application to;
- Configure counting direction
- Configure position of the entry and exit area
- Configure calibration of the people size
- Configure algorithm parameters
- Configure Cloud service url
- Enable/Disable WiFi
- Imaging Sensor configuration
- Firmware upgrade
- Update installation configuration to on board flash
Note: Please note that WiFi connectivity is provided as an example only, customers are expected to modify/replace this based on their application.
People Counter on BLIP is an ideal solution to get data related to people entering and exiting a specified area that can be used to drive several business analytics. People Counter product counts the number of entries and exits of people in a specified area based on the direction of movement. By counting and reporting the number of people by this product, when coupled with powerful business analytics can enable multiple applications such as retail analytics, footfall measurement in malls, counting at building entrances and zones of Interest. The product operates in real time, fully embedded to run on edge in the BLIP. The various connectivity options on SPI/UART/WiFi provides users with options to couple it with multiple end systems and also enable services on cloud. The intuitive user interface makes it easy to use, install and configure the system.
Each module supports the Analog Devices, Inc. (ADI) Blackfin or SHARC Processor family and is a licensed product that is available in object code format. Recipients must sign or accept a license agreement with ADI prior to being shipped or downloading the modules identified in the license agreement.
|Software||Code (KB)||Data (MB)||MIPS|
|People Counter Suite||64.94||1.032||230|
- MIPS measured on optimal memory layout, VGA input running on an ADSP-BF707-BLIP board @ 30 fps.
- Measurements done with CCLK1=384, SCLK2=192 for ADSP-BF707
- Code compatible across all ADSP-BF70x with silicon anomaly workarounds implemented based on ADSP-BF707 Silicon Revision 0.0
1CCLK - Core clock
2SCLK - System Clock
|Low Density Scene||0.91||0.94||0.93|
|Medium Density Scene||0.83||0.87||0.85|
Specificity shows how good or specific the algorithm is at detecting actual humans while ignoring false alarms.
Sensitivity shows how sensitive the algorithm is at detecting humans and not missing them.
Low Density Scenario – The density of people is around 2-3 people crossing the field of view at a given time
Medium Density Scenario – The density of people is around 4-5 people crossing the field of view at a given time
- Windows 7 Professional/Enterprise/Ultimate (32 and 64 bit). It is recommended to install the software in a non-UAC-protected location.
- Minimum of 2 GHz single core processor, 3.3 GHz dual core is recommended.
- Minimum of 4 GB memory (RAM).
- Minimum of 2 GB hard disk (HDD) space is required.
- CrossCore Embedded Studio for Analog Devices Processors.