← Back to selected work
CASE / 02Private archive

YOLO KZ

A field computer-vision system that turns a tripod-mounted phone camera into directional counts for vehicles, people, bicycles and scooters.

  • YOLO
  • Computer vision
  • C#
  • ASP.NET Core
  • PostgreSQL
  • Docker
SYS/02
  1. 01Detect object
  2. 02Cross line
  3. 03Record count
SYSTEM MAP / 02

Signal path

The working path from input to physical or measurable outcome.

FIELD SENSORPhone / tripod
INFERENCEYOLO tracker
INGESTIONASP.NET Core
ANALYTICSPostgreSQL
↳ OUTPUTDirectional counts

Car · person · bicycle · scooter

RoleComputer vision & backend architecture
Year2026
Field sensorPhone camera / tripod
ClassesCar / person / bike / scooter
OutputDirectional live counts
01

The challenge

Start with the constraint.

Street activity had to be measured from a portable camera setup and converted into reliable, class-specific directional statistics.

02

The approach

Make complexity local.

  1. 01

    Used a phone on a tripod as a portable field camera for repeatable observation points.

  2. 02

    Detected and classified cars, pedestrians, bicycles and scooters with a YOLO pipeline.

  3. 03

    Converted tracked crossings into directional events and idempotent offline batches.

  4. 04

    Built the ASP.NET Core backend around projects, devices, protected analytics and live counters.

03

The outcome

A portable traffic-observation workflow with live category counts and a backend prepared for controlled customer analytics.