Self-Hosted C++ Server

Computer Vision & Media Processing

A high-performance C++ server with 20+ image, video, text, and NLP processing capabilities. Connect via TCP JSON protocol or MCP tools for seamless AI integration.

Get Started Read the Docs
20+
Capabilities
C++17
Performance
TCP
JSON Protocol
inigami-server
$ curl -X POST localhost:10210/process 
// Request
{
  "type": "detect_objects",
  "image_path": "photo.jpg",
  "request_id": "abc-123"
}

// Response
{
  "request_id": "abc-123",
  "result": {
    "objects": [
      {
        "class": "person",
        "confidence": 0.95,
        "bbox": { "x": 120, "y": 80 }
      }
    ]
  }
}

Core Capabilities

From object detection to NLP, Inigami provides a unified server for all your media processing needs.

🎯

Object Detection

YOLOv5-powered object detection via OpenCV DNN. Identify and locate objects in images with bounding boxes, class labels, and confidence scores.

🌊

Depth Estimation

Monocular depth map generation using edge detection and Sobel gradients. Produces JET-colormapped depth visualizations from single images.

🧩

Image Segmentation

K-means clustering in LAB color space for automatic image segmentation. Generates distinct segment masks for scene understanding.

🖼️

Image Transforms

Resize, crop, rotate, flip, and apply artistic filters powered by ImageMagick. Chain multiple transformations in a single request.

📝

Text Analysis

Word count, character statistics, spell checking, readability scoring, and text extraction. Process documents and strings with precision.

🧠

NLP Pipeline

Natural language processing with tokenization, sentiment analysis, entity extraction, and language detection. Understand text at scale.

View All 20+ Capabilities

How It Works

Inigami uses a simple TCP JSON protocol. Send a request, get structured results. No complex SDKs required.

01

Send Request

Your application sends a JSON request over TCP to the Inigami server specifying the processing operation and input data.

02

Process

The C++ server routes the request to the appropriate processing pipeline -- OpenCV, ImageMagick, or the NLP engine.

03

Receive Results

Structured JSON results are returned including processed data, file paths, detection results, or analysis output.

TCP JSON Protocol
// Request -- detect objects in an image 
{
  "request_id": "req-001",
  "type": "detect_objects",
  "image_path": "/data/photo.jpg"
}
// Response -- structured detection results 
{
  "request_id": "req-001",
  "result": {
    "objects": [
      {
        "class": "person",
        "confidence": 0.95,
        "bbox": { "x": 120, "y": 80, "w": 200, "h": 350 }
      },
      {
        "class": "car",
        "confidence": 0.88,
        "bbox": { "x": 400, "y": 200, "w": 300, "h": 180 }
      }
    ]
  }
}
Learn More About the Architecture

Ready to Process Media at Scale?

Get Inigami running in minutes. Build, start the server, and connect with TCP JSON or MCP tools.

Get Started Now