OMSPEC is a high-performance, hardware-aware orchestration engine designed for large-scale multispectral image processing. It is engineered to handle 1TB+ datasets by utilizing a native D-language scheduler that manages parallel Python worker processes.

OMSPEC uses a native D Master Process to manage worker lifecycles. This ensures minimal overhead and maximum CPU utilization during alignment tasks.

The scheduler monitors system memory usage in real-time. It dynamically pauses the spawning of new workers if memory usage exceeds the user-defined threshold in omspec.cfg, preventing system swap-lag or out-of-memory crashes.

Each task is tracked via a Process Control Block (PCB). If a worker fails, the master process automatically retries the task (configurable) before flagging it, ensuring long-running jobs are resilient against transient errors.

OMSPEC ships with an embedded Python 3.11 interpreter and all necessary engine scripts. No global Python installation or pip install is required on the target machine. It just-works!

OMSPEC_v1.0.0/
├── bin/
│   ├── omspec.exe      # Master Orchestrator Binary
│   └── omspec.cfg      # Hardware & Planning Configuration
├── engine/
│   └── *.py            # Engine scripts for task processing
├── python_3_11_14/     # Embedded Python runtime
└── LICENSE.txt

Users can tune OMSPEC to their hardware and workflow:

• max_memory_mb: Maximum RAM usage in MB (e.g., 8192 for 8GB). Scheduler pauses spawning workers above this limit.
• max_retries: Number of times a failed chunk is retried (default: 3).
• tick_interval_ms: Frequency of the scheduler monitor loop in milliseconds.
• bands: Comma-separated list of expected spectral bands (e.g., r, g, b, nir, re).

v1.0.0 (Current)

• High-speed parallel orchestration logic.
• Runtime hardware resource accounting.
• Mock-mode validation for pipeline I/O.

v1.1.0 (Upcoming)

• Graceful termination with SIGINT (Ctrl+C) handling.
• Real-time telemetry dashboard showing RAM usage and worker statuses.

v1.2.0+

• Production-grade alignment, tiling, and testing services.
• Support for Linux

v2.0

• Cloud integration
• Extensibility with Custom Image Processing Pipelines

1. Extract the ZIP archive.
2. Navigate to the bin/ directory (or add it to PATH to use anywhere)
3. Run the tool:

omspec --target=<your_target_directory> [options]

Logs are generated in the log/ directory (or current working directory in v1.0.0) detailing planning and execution phases.

Please edit your omspec.cfg file for higher level control of your OMSPEC tool.

Example:

omspec.exe --align --target=D:\Datasets\MyProject --depth=5

OMSPEC is released under the Polyform Non Commercial License. See LICENSE.txt for full terms.

• Currently, log files are generated in the current working directory. In future versions, logs will be written to the target directory.
• Engine scripts are editable but modifying them may break the workflow. Use caution when editing engine/*.py.