OpenZL

An open-source library for computer vision

Posted by Joshua Z. Zhang on January 20, 2015

A minimal cross-platform library for computer vision projects with ability to read/write/display images.

Github repository

OpenZL

You may want to try this new repository with more functions and more elegant way to start writing code.

Introduction

OpenZL is a light weighted portable X-platform library packed with fundamental functions for simple Computer Vision projects.

Wrapped in C++ as itself lacks image IO libraries, meanwhile provide more convenience over C.

Full documentation in doc/html/

License

MIT License.

Usage

  • Make sure you have c++11 support(vc11+, -std=c++11 in gcc 4.6+/clang3.3+)

  • include “core.hpp” for basic class

  • include “imgcodecs.h” for image read/write capabilities

  • include “highgui.h” for GUI related functions, e.g. imshow().

  • using namespace zl

Examples

Check out README.

Capability

  • Get rid of bulky OpenCV library and messy configuration steps when you want start a very simple app.

  • Classes like Point, Rect, Mat are implemented either copied from OpenCV or rewritten to make it simple and clean.

  • Support cross-platform time() function.

  • Support cross-platform waitkey() function without pressing enter.

  • Support read image from BMP,JPG,PNG, TGA, PSD, GIF, HDR, PIC (thanks to Sean T. Barrett’s stb_image library).

  • Support save image to BMP, JPG, PNG, TGA (thanks to Sean T. Barrett’s stb_image_write library and Jon Olick’s jpeg writer).

Acknowledgement

stb_image Library by Sean T. Barrett JPEG_Writer by Jon Olick

  • Support image display (thanks to CImg library).

  • Support various drawing functions(line, rectangle, circle, polygon…).