# ranker-adhipkumar/ai-image-restoration architecture

> Architecture diagram of the GitHub repository ranker-adhipkumar/ai-image-restoration, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-02). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/ranker-adhipkumar/ai-image-restoration
- Repository: https://github.com/ranker-adhipkumar/ai-image-restoration

## Overview

This is a CPU-friendly image-restoration application with Gradio and CLI entry points. Users can simulate damage for benchmarking or submit already-damaged images for blind diagnosis and restoration; the system then presents restoration results and quality assessments. The graph preserves the documented restoration modes, model fallbacks, sample-image support, and comparison visualizations. Some subsystem links come from the SOURCE INDEX or README because their implementation was not excerpted; those limits are reflected by omitting uncertain links.

Main entry points: [`app.py`](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/app.py) (Gradio web app), [`restoration/__init__.py`](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/__init__.py) (Restoration dispatcher), [`main.py`](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/main.py) (Command-line app), [`restoration/blind_restorer.py`](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/blind_restorer.py) (Blind restoration)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_entry["User interfaces"]
  node_web["Gradio web app<br/>[app.py]"]
  node_cli["Command-line app<br/>[main.py]"]
end

subgraph group_workflow["Image workflows"]
  node_simulation["Degradation simulation<br/>[degradation.py]"]
  node_blind["Blind restoration<br/>[blind_restorer.py]"]
end

subgraph group_restoration["Restoration engines"]
  node_dispatch["Restoration dispatcher<br/>[__init__.py]"]
  node_denoise["Denoising<br/>[denoiser.py]"]
  node_deblur["Deblurring<br/>[deblurrer.py]"]
  node_inpaint["Inpainting<br/>[inpainter.py]"]
  node_artifacts["Artifact removal"]
  node_nafnet["NAFNet architecture<br/>[nafnet_arch.py]"]
end

subgraph group_quality["Quality and reporting"]
  node_diagnostics["Blind diagnostics<br/>[diagnostics.py]"]
  node_metrics["Reference metrics<br/>[metrics.py]"]
  node_visualizer["Comparison visualizer<br/>[visualizer.py]"]
  node_demo["Benchmark output generator"]
end

subgraph group_support["Image support"]
  node_utils["Image I/O and helpers<br/>[utils.py]"]
  node_samples["Sample downloader"]
end

node_user(("Image user"))

node_user -->|"uploads image"| node_web
node_user -->|"provides image"| node_cli
node_web -->|"requests degradation"| node_simulation
node_web -->|"requests restoration"| node_blind
node_web -->|"requests restoration"| node_dispatch
node_web -->|"requests diagnosis"| node_diagnostics
node_web -->|"requests metrics"| node_metrics
node_web -->|"renders comparisons"| node_visualizer
node_web -->|"uses sample images"| node_samples
node_cli -->|"requests degradation"| node_simulation
node_cli -->|"requests restoration"| node_blind
node_cli -->|"requests restoration"| node_dispatch
node_cli -->|"requests diagnosis"| node_diagnostics
node_cli -->|"requests metrics"| node_metrics
node_dispatch -->|"dispatches noise modes"| node_denoise
node_dispatch -->|"dispatches blur modes"| node_deblur
node_dispatch -->|"dispatches inpainting"| node_inpaint
node_dispatch -->|"dispatches artifacts"| node_artifacts
node_dispatch -->|"dispatches blind modes"| node_blind
node_blind -->|"diagnoses image"| node_diagnostics
node_blind -->|"applies denoising"| node_denoise
node_blind -->|"applies deblurring"| node_deblur
node_denoise -->|"uses model"| node_nafnet
node_deblur -->|"uses model"| node_nafnet
node_metrics -->|"uses diagnostics"| node_diagnostics
node_demo -->|"uses degradation"| node_simulation
node_demo -->|"uses restoration"| node_dispatch
node_demo -->|"computes metrics"| node_metrics
node_demo -->|"creates comparisons"| node_visualizer
node_web -->|"uses image helpers"| node_utils
node_cli -->|"loads and resizes"| node_utils
node_simulation -->|"uses image helpers"| node_utils
node_diagnostics -->|"uses image helpers"| node_utils
node_metrics -->|"uses image helpers"| node_utils
node_visualizer -->|"uses image helpers"| node_utils

click node_web "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/app.py"
click node_cli "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/main.py"
click node_simulation "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/degradation.py"
click node_blind "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/blind_restorer.py"
click node_dispatch "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/__init__.py"
click node_denoise "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/denoiser.py"
click node_deblur "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/deblurrer.py"
click node_inpaint "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/inpainter.py"
click node_artifacts "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/artifact_remover.py"
click node_nafnet "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/nafnet_arch.py"
click node_diagnostics "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/diagnostics.py"
click node_metrics "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/metrics.py"
click node_visualizer "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/visualizer.py"
click node_demo "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/generate_demo_outputs.py"
click node_utils "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/utils.py"
click node_samples "https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/sample_images/download_samples.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_web,node_cli,node_user toneBlue
class node_simulation,node_blind toneAmber
class node_dispatch,node_denoise,node_deblur,node_inpaint,node_artifacts,node_nafnet toneMint
class node_diagnostics,node_metrics,node_visualizer,node_demo toneRose
class node_utils,node_samples toneIndigo
```

## Components

### User interfaces

- [Gradio web app](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/app.py): `app.py`
- [Command-line app](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/main.py): `main.py`

### Image workflows

- [Degradation simulation](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/degradation.py): `degradation.py`
- [Blind restoration](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/blind_restorer.py): `restoration/blind_restorer.py`

### Restoration engines

- [Restoration dispatcher](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/__init__.py): `restoration/__init__.py`
- [Denoising](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/denoiser.py): `restoration/denoiser.py`
- [Deblurring](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/deblurrer.py): `restoration/deblurrer.py`
- [Inpainting](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/inpainter.py): `restoration/inpainter.py`
- [Artifact removal](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/artifact_remover.py): `restoration/artifact_remover.py`
- [NAFNet architecture](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/restoration/nafnet_arch.py): `restoration/nafnet_arch.py`

### Quality and reporting

- [Blind diagnostics](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/diagnostics.py): `diagnostics.py`
- [Reference metrics](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/metrics.py): `metrics.py`
- [Comparison visualizer](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/visualizer.py): `visualizer.py`
- [Benchmark output generator](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/generate_demo_outputs.py): `generate_demo_outputs.py`

### Image support

- [Image I/O and helpers](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/utils.py): `utils.py`
- [Sample downloader](https://github.com/ranker-adhipkumar/ai-image-restoration/blob/main/sample_images/download_samples.py): `sample_images/download_samples.py`

### Other components

- **Image user**

## Connections

- Image user → Gradio web app: uploads image (evidence: `README.md`)
- Image user → Command-line app: provides image (evidence: `README.md`)
- Gradio web app → Degradation simulation: requests degradation (evidence: `app.py`)
- Gradio web app → Blind restoration: requests restoration (evidence: `app.py`)
- Gradio web app → Restoration dispatcher: requests restoration (evidence: `app.py`)
- Gradio web app → Blind diagnostics: requests diagnosis (evidence: `app.py`)
- Gradio web app → Reference metrics: requests metrics (evidence: `app.py`)
- Gradio web app → Comparison visualizer: renders comparisons (evidence: `app.py`)
- Gradio web app → Sample downloader: uses sample images (evidence: `app.py`)
- Command-line app → Degradation simulation: requests degradation (evidence: `main.py`)
- Command-line app → Blind restoration: requests restoration (evidence: `main.py`)
- Command-line app → Restoration dispatcher: requests restoration (evidence: `main.py`)
- Command-line app → Blind diagnostics: requests diagnosis (evidence: `main.py`)
- Command-line app → Reference metrics: requests metrics (evidence: `main.py`)
- Restoration dispatcher → Denoising: dispatches noise modes (evidence: `restoration/__init__.py`)
- Restoration dispatcher → Deblurring: dispatches blur modes (evidence: `restoration/__init__.py`)
- Restoration dispatcher → Inpainting: dispatches inpainting (evidence: `restoration/__init__.py`)
- Restoration dispatcher → Artifact removal: dispatches artifacts (evidence: `restoration/__init__.py`)
- Restoration dispatcher → Blind restoration: dispatches blind modes (evidence: `restoration/__init__.py`)
- Blind restoration → Blind diagnostics: diagnoses image (evidence: `restoration/blind_restorer.py`)
- Blind restoration → Denoising: applies denoising (evidence: `restoration/blind_restorer.py`)
- Blind restoration → Deblurring: applies deblurring (evidence: `restoration/blind_restorer.py`)
- Denoising → NAFNet architecture: uses model (evidence: `restoration/denoiser.py`)
- Deblurring → NAFNet architecture: uses model (evidence: `restoration/deblurrer.py`)
- Reference metrics → Blind diagnostics: uses diagnostics (evidence: `metrics.py`)
- Benchmark output generator → Degradation simulation: uses degradation (evidence: `generate_demo_outputs.py`)
- Benchmark output generator → Restoration dispatcher: uses restoration (evidence: `generate_demo_outputs.py`)
- Benchmark output generator → Reference metrics: computes metrics (evidence: `generate_demo_outputs.py`)
- Benchmark output generator → Comparison visualizer: creates comparisons (evidence: `generate_demo_outputs.py`)
- Gradio web app → Image I/O and helpers: uses image helpers (evidence: `app.py`)
- Command-line app → Image I/O and helpers: loads and resizes (evidence: `main.py`)
- Degradation simulation → Image I/O and helpers: uses image helpers (evidence: `degradation.py`)
- Blind diagnostics → Image I/O and helpers: uses image helpers (evidence: `diagnostics.py`)
- Reference metrics → Image I/O and helpers: uses image helpers (evidence: `metrics.py`)
- Comparison visualizer → Image I/O and helpers: uses image helpers (evidence: `visualizer.py`)
