A web-based educational tool for teaching visual composition and prompt engineering for AI-assisted multi-layered image assembly, using a non-destructive, layer-based approach.
The Photo Composition Prompt Generator teaches students how to structure complex prompts for AI image generation models by decomposing multi-layered image assembly into discrete, explicit operations. Rather than writing vague requests like "combine these photos," students learn to specify each composition decision through precise parameters — subject isolation, spatial positioning, scaling, blend modes, edge control, and color adjustments.
The tool operates on a non-destructive, layer-based paradigm: uploaded images are treated as distinct layers that are composed through explicit instructions, never altered or regenerated by the AI.
- Multi-Image Composition — Upload 2-3 images and define their roles (background, overlay, subject)
- Layer-Based Workflow — Non-destructive composition using the metaphor of Photoshop layers
- Role Assignment — Define each image as main subject, background, or overlay element
- Blend Modes — Normal, Multiply, Screen, Overlay, Soft Light, Hard Light, and more
- Spatial Controls — Position (X/Y), scale, rotation, and opacity for each layer
- Subject Cutout — Optional background removal before compositing
- Edge Control — Feathering, masking, and shadow generation parameters
- Structured Prompt Output — Generates detailed, machine-readable composition instructions
- Final Output Settings — Resolution, format, and quality specifications
| File | Description |
|---|---|
| index.html | Landing page |
| photo-composition-generator_APP.html | Main application |
| photo-composition-prompt-learning-manual.html | Educational Manual (12 chapters) |
https://gkorakakis.github.io/photo-composition-prompt-generator/
Open directly in any modern browser. No installation required.
- Download or clone this repository
- Open index.html in any modern web browser
- Upload 2-3 images (JPG/PNG) and define their composition roles
- No server, no installation, no dependencies required
Developed for the Interactive Design course at the Department of Graphic Design and Visual Communication, University of West Attica (UNIWA).
Students learn to:
- Decompose complex visual compositions into discrete layer operations
- Specify spatial relationships, blend modes, and masking parameters explicitly
- Translate intuitive compositional ideas into structured, machine-readable prompts
- Understand the relationship between traditional compositing and AI-mediated workflows
- Architecture: Single-file HTML/CSS/JavaScript (no framework dependencies)
- Image Handling: Client-side FileReader API, no server upload
- Browser Support: Chrome, Firefox, Safari, Edge (latest versions)
- Responsive: Adaptive layout for desktop and tablet
Georgios Korakakis Assistant Professor, Department of Graphic Design and Visual Communication School of Applied Arts and Culture, University of West Attica (UNIWA) gkorakakis@uniwa.gr
This software was developed with the assistance of Claude AI (Anthropic) as a coding aid. All pedagogical design, scientific content, composition methodology, and educational framework were conceived, validated, and supervised by the author.
Korakakis, G. (2026). Photo Composition Prompt Generator: An Educational Tool for AI-Assisted Multi-Layered Image Composition (v1.0) [Software]. Zenodo. https://doi.org/10.5281/zenodo.19474264
Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) https://creativecommons.org/licenses/by-nc/4.0/
https://doi.org/10.5281/zenodo.19474264
Department of Graphic Design and Visual Communication - School of Applied Arts and Culture - University of West Attica (UNIWA)