About AI-Assisted AesthetiQR

Inspiration: While working on marketing and design projects, I noticed that QR codes often look visually jarring and disrupt the aesthetics of posters, flyers, and product packaging. I wanted to create a tool that makes QR codes context-aware, visually harmonious, and scannable, blending seamlessly with any design.

What I Learned: During this project, I deepened my knowledge in computer vision, image processing, and machine learning, including KMeans clustering for color extraction and edge/saliency detection for safe placements. I also explored AI prompt generation for design automation, and how to combine technical and aesthetic considerations in a single solution.

How I Built It: Image Analysis: Uploaded images are processed to extract dominant colors and detect textures using Laplacian variance.

Safe Zone Detection: Edges and saliency are detected to identify regions free of text or logos where QR codes can be placed.

AI Design Prompt Generation: Using extracted colors, texture classification, and safe zones, the system generates a detailed AI-ready prompt for designing an aesthetic QR code.

Tools & Tech Stack: Python, OpenCV, scikit-learn, Webcolors, Matplotlib, Pillow, Google Colab.

Challenges Faced: Ensuring accurate color detection on diverse images with varying lighting and contrast. Designing robust safe-zone detection that avoids text, logos, and busy areas without being too restrictive. Balancing aesthetic appeal vs. QR scan ability, especially for highly textured or complex backgrounds. Translating technical image data into human-readable AI prompts that are both descriptive and effective.

Outcome: The project successfully generates AI-ready, aesthetically pleasing QR codes tailored to the uploaded image, making it suitable for marketing, branding, and digital art applications. It’s a perfect blend of AI, design, and functionality.

Built With

Share this project:

Updates