🖼️ InnoVarsity: Poster AI Evaluator Module

📌 Overview

This repository contains the Poster Analysis Module for InnoVarsity (an AI-driven Science Fair Online Exhibition platform).

Traditional academic posters are usually evaluated manually, which is prone to human bias and inconsistency. This Python-based AI module autonomously evaluates research posters (Panaflex) out of 30 Marks based on three core dimensions: Visual Aesthetics, Structural Layout, and Academic Depth.

📌 Output

🧠 Core Evaluation Metrics (The 3 Dimensions)

1. Aesthetics & UI (10 Marks):

   • Blur Detection: Uses cv2.Laplacian variance to ensure the poster is sharp and readable.
   • Contrast Analysis: Uses Standard Deviation (np.std) to check the color harmony between text and background.
   • Information Crowding: Uses Canny Edge Density to ensure a perfect balance between text, diagrams, and whitespace.

2. Structural Layout (10 Marks):

   • Instead of basic color thresholding, this module uses Canny Edge Detection combined with Morphological Transformations (Dilation/Closing).
   • It dynamically calculates the area of the poster and counts distinct academic blocks (e.g., Abstract, Methodology, Results) regardless of the background color or gradient.

3. Academic Depth & NLP (10 Marks):

   • Uses Tesseract OCR to extract raw text from the image.
   • Calculates Lexical Density (ratio of unique words to total words) to evaluate the complexity of the English used.
   • Matches extracted text against a custom Computer Science/Engineering vocabulary dictionary to reject irrelevant or fake images.

📂 Project Structure

POSTER_AI_MODULE/
│
├── poster_evaluator.py      # The main AI evaluation class and logic
├── sample_poster.jpg        # Sample academic poster for testing
├── poster3.jpg              # Additional test image
└── README.md                # Project documentation