computer-vision

Simple image-on-image search project.

Table of Contents

• Technologies Used
• Features
• Screenshots
• Setup
• Usage
• Project Status
• Room for Improvement
• License

Technologies Used

• OpenCV - version 4.5.4
• fmt - version 8
• stringr - version 1.4.1
• Rcpp - version 1.0.11

Features

• Crossplatform ( X11 based desktop environments, Windows ).
• Reading image from both file or window.
• Mouse clicks and movement.

Screenshots

• Template: Sample:

• Template: Sample:

• Template: Sample:

Setup

• OpenCV for Debian/ Ubuntu:

  apt install libopencv-dev

  Other platforms: OpenCV Releases

• fmt library for Debian/ Ubuntu:

  apt install libfmt-dev

  Other platforms: fmt library

• R stringr package:

  install.packages("stringr")

• R Rcpp package:

  install.packages("Rcpp")

Usage

Change window_name in src/main.r. Put images into ./image folder. Image name example: for sample is "sample.category_name.png" for template is "template.category_name.unique_name.png" Run ./run.sh

The image you are looking for should have the same size on sample as on template.

Project Status

Project is: in progress.

Room for Improvement

Room for improvement:

• Improve RAM usage.
• Make search for template independent from sample scale.

To do:

• Add GPU as an option.
• Add regular expression support for Windows.

License

This project is open source and available under the GNU General Public License v3.0.