Farmtrace facial recognition

Facial recognition that works where the work happens.

Identify registered employees directly on the device—even when connectivity, lighting, camera position, and crowding are less than perfect.

How recognition works

Local matching after synchronization · Guided quality checks · Liveness protection

Every scan follows the same clear sequence.

Capture a usable frame

Quality gates guide the operator with clear prompts such as move closer, hold still, or find more light.
  • Face size and pose
  • Brightness and stability
  • Single-subject check

Understand the face

The face is aligned and checked for signs that it belongs to a real person before it is compared.
  • Consistent positioning
  • Liveness protection
  • Recognition profile

Search nearby candidates

The device quickly compares the captured face with enrolled people nearby in the local recognition set.
  • Fast candidate search
  • Local recognition
  • Responsive at the point of work

Accept or refuse the decision

A weak result or an ambiguous runner-up becomes unknown instead of being forced into an identity.
  • Clear acceptance rules
  • Ambiguity checks
  • Unknown when uncertain

Recognition built for real attendance and workforce operations.

Recognition happens locally
After synchronization, enrolled employee profiles remain available on the device without depending on live connectivity.
Better input, safer recognition
Clear guidance helps employees correct lighting, positioning, movement, or competing faces before matching begins.
Unknown is a valid answer
FarmTrace returns an identity only when the leading result is clear enough to trust.
Built for busy entrances
The workflow can focus on the dominant subject while refusing a scan when another face makes the result ambiguous.

Choose the right balance between easier capture and stronger verification.

Start with Balanced, move toward easier field capture or stronger controlled-access verification, or choose a custom implementation.

01

Lenient

For variable lighting, outdoor work, and devices that need easier capture.

02

Moderate

Two-frame verification with practical tolerance for changing everyday conditions.

03

BalancedRecommended

Recommended for most teams: dependable two-frame verification and quality control.

04

Strict

Three-frame verification for controlled access points with one person in view.

05

Maximum

Four-frame verification for high-security workflows and tightly controlled capture.

06

Custom implementation

Tailored capture and verification rules for workflows beyond the presets.

Connected to FarmTraceFrom one employee photo to recognition in Dsync.

Send one approved employee photo to the FarmTrace API. It generates the face embedding and synchronizes the recognition profile to Dsync—without capturing the same person on every device.

1

Enrol from one photo

Use an approved employee image already available to your business.

2

Generate the recognition profile

The FarmTrace API prepares the face embedding for recognition.

3

Put it to work in Dsync

Synchronize the profile to supported devices where teams clock in.

Prepare each site for dependable recognition.

Set up the people, operating profile, and devices before teams begin scanning.

01.
Enrol employees
Create employee profiles centrally from approved images through the FarmTrace API.
02.
Choose the operating profile
Start with Balanced or tune verification around the site’s lighting, traffic, and access requirements.
03.
Prepare supported devices
Confirm camera, location, battery, storage, and connectivity before rollout.

Ready for the devices your team depends on.

Choose Android hardware that can keep recognition responsive through a full day of field work.

Operating system
Android 7.0 (Nougat) or newer
Memory
6 GB RAM or more for responsive day-to-day use
Storage
At least 8 GB free, with additional capacity for cached field data
Processing
Qualcomm octa-core 2.45 GHz processor or newer
Connectivity
Stable 4G/LTE or Wi-Fi with at least 2 Mbps upload and download
Cameras
5 MP front camera and 8 MP rear camera, or better
Location
Enabled, reliable GPS for accurate field records
Power
3000 mAh battery or higher for full-day operation

* For dependable high-volume scanning, we recommend the DT50S with a charging dock, available directly through FarmTrace. If you source another device, confirm its GPS accuracy, processing performance, memory capacity, and battery life before rollout; FarmTrace cannot warrant the performance of independently selected hardware.

Download supported devices (PDF)