Welcome to the Ovative Group Image Caption Generator’s documentation!

This project was created by NDSU’s 2024 Capstone class. Contributors are: Ian King, Omar Elsegeiny, Mason Myles, Bohdan Ivanyshyn, and Emily Anderson.

The Problem

Ovative Group is a digital media and marketing firm that provides analysis of their clients’ ads. However, in order to analyze the ads, they need to describe them. This is often time intensive. They came to us with the request to design an Image Captioning program that would allow the advertisements to be analyzed easier.

Our Solution

A command-line interface that intakes one or multiple images from a URL(s) or a directory. It then runs the images through several models and returns those captions in a CSV file. This CSV file, once created, then compiles all future captions in one place to allow for easier analysis.

Contents:

• Overview
  • Project Overview: Ovative Group Capstone
  • Features
  • Output
  • Efficiency
• Sales Force Model
  • caption_image Function
  • Example usage:
• Coherence Model
  • Coherence Function
  • Guidelines:
  • Example usage:
• Easy OCR Model
  • download_image Function
  • draw_boxes Function
  • inference Function
  • Example usage:
• App Module
  • Input Manager
  • Image Processor
  • Storing Outputs
• Threading
• Utility Functions
  • clear()
  • quitting(user_input: str) -> bool
  • end() -> None
  • break_line()
  • is_dir(directory: str) -> bool
  • is_url(image_url: str) -> bool
  • clean_url(user_input: str) -> str
  • create_dir() -> None
  • output_file_path() -> str
  • validate_and_add_url(img_source: str, image_sources: list[str]) -> None
  • validate_and_add_files(directory: str, image_sources: list[str]) -> None
  • is_valid_extension(img_source: str) -> bool
  • filter_and_validate(image_sources: list[str]) -> list[str]
• Sources