Technologies and engineering approach

XRayDetect operates where three disciplines meet: image processing, embedded systems and system integration. The technologies listed below are the foundation on which we engineer our X-ray system solutions.

Software

Area Technologies
Image-processing core C++ and Rust, multithreading, SIMD optimization
Visualization OpenGL (programmable pipeline), GPU hardware acceleration
Operator UI Qt (Widgets / QML)
Plug-in modules DLL / SO exposing a C-compatible ABI (extern "C")
Networking TCP/UDP, detector and controller protocols
Platforms Windows 10/11 (64-bit), Linux (Ubuntu 20.04+ and equivalents)
Object recognition / AIClassical computer vision plus deep learning (CNN), real-time inference, on-prem/edge deployment

Principles:

  • real-time processing, with the heaviest operations moved onto dedicated worker threads;
  • thread-safe context objects that let several images be processed in parallel;
  • a modular design — algorithms such as colorization can be added without rewriting the core software;
  • backward-compatible API versioning (semver).

Embedded systems

Area Technologies
Microcontrollers 32-bit ARM Cortex-M (and equivalents)
Firmware C and Rust (bare-metal / RTOS)
Communication on-chip Ethernet 10/100, TCP/UDP
Peripherals relay outputs, 24 V discrete inputs, ADC, displays
Reliability watchdog, galvanic isolation, EM-interference protection
Updates over-the-network remote firmware updates

Principles:

  • a guaranteed reaction time for critical events such as emergency stop;
  • built-in self-diagnostics and health monitoring;
  • a design intended for serial production and round-the-clock operation.

Hardware integration

  • Detectors: industrial line-scan detectors — vendor SDKs, acquisition protocols and high-speed camera interfaces.
  • Synchronization: encoders, barrier sensors and trigger modes.
  • X-ray generators: on/off control, safety timers and error handling.
  • Mechanics: conveyors, carriages and interlock systems.

Engineering practices

  • Version control with a well-structured repository layout.
  • Mock implementations — emulated interfaces that let teams work in parallel before the real hardware is available.
  • Documentation of the API and architecture, delivered as part of the project.
  • Testing on actual hardware: calibration, load and acceptance testing.
  • Source code and rights handed over by default.

Related sections

  • Services — how these technologies are applied in practice.
  • Projects — examples of work we have delivered.
Contact

Let's talk about your project

We'll walk through the technical approach to your project together, assess the system and propose a fitting solution architecture.