Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/codeplaysoftware/visioncpp

A machine vision library written in SYCL and C++ that shows performance-portable implementation of graph algorithms
https://github.com/codeplaysoftware/visioncpp

sycl

Last synced: 2 months ago
JSON representation

A machine vision library written in SYCL and C++ that shows performance-portable implementation of graph algorithms

Awesome Lists containing this project

README 

        


  






# Overview

**VisionCpp** is a lightweight header-only library for computer vision and image processing.

The aim of the library is to provide a toolbox that enables performance portability for heterogeneous platforms using modern C++.



Written using [SYCL 1.2.1](https://www.khronos.org/registry/SYCL/specs/sycl-1.2.1.pdf) and compiled/tested with [ComputeCpp](https://codeplay.com/products/computesuite/computecpp) to accelerate vision code using OpenCL devices.



## Table of contents

* [Integration](#integration)

* [Tutorials](#visioncpp-tutorials)

* [Sample Code](#sample-code)

* [Requirements](#requirements)

* [Build](#build)

* [Examples](#examples)

* [Documentation](#documentation)

* [Contributing](#contributing)

* [Resources](#resources)

* [License](#license)

* [Known Issues](#known-issues)



## Integration

You will need to install ComputeCpp in order to use VisionCpp, you can follow the [ComputeCpp Getting Started guide](https://developer.codeplay.com/products/computecpp/ce/guides/getting-started) that outlines the installation process.

All you need to do is include the VisionCpp.hpp header in your project and you are good to go! ( assuming that OpenCL and ComputeCPP is installed correctly. )



~~~~~~~~~~~~~~~{.cpp}

#include  //all that is needed

~~~~~~~~~~~~~~~



## VisionCpp Tutorials

There are some tutorials explaining how to perform different operations using VisionCpp. These cover basic [Hello World](https://github.com/codeplaysoftware/visioncpp/wiki/Example:-Hello-World), [Anisotropic Diffusion](https://github.com/codeplaysoftware/visioncpp/wiki/Example:-Anisotropic-Diffusion), [Bayer Filter Demosaic](https://github.com/codeplaysoftware/visioncpp/wiki/Example:-Bayer-Filter-Demosaic), [Dense Depth Reconstruction with Block Matching Algorithm](https://github.com/codeplaysoftware/visioncpp/wiki/Example:-Dense-Depth-Reconstruction-with-Block-Matching-Algorithm) and [Harris Corner Detection](https://github.com/codeplaysoftware/visioncpp/wiki/Example:-Harris-Corner-Detection).



## Sample Code

Below is a very simple application that will do the conversion RGB -> HSV. Full source code can be found in the examples folder.

RGB is assumed to be a three-channel unsigned char storage with a reasonable channel order.



~~~~~~~~~~~~~~~{.cpp}

  // main, args, checks and all the boring stuff



  // ...



  // where VisionCpp will run.

  auto dev = visioncpp::make_device();



  // create a host container for input data

  std::shared_ptr in_rgb(new unsigned char[3],

  [](unsigned char *dataMem) { delete[] dataMem;});



  in_rgb.get()[0] = atoi(argv[1]);

  in_rgb.get()[1] = atoi(argv[2]);

  in_rgb.get()[2] = atoi(argv[3]);



  // create a host container for output data

  std::shared_ptr out_hsv(new unsigned char[3],

  [](unsigned char *dataMem) { delete[] dataMem;});



  // exiting this scope will sync data

  {

    // definition of the VisionCpp pipeline:



    // create terminal nodes - a leaf node ( data node ) of the expression tree.

    // terminal struct takes 4 arguments

    // 1st template parameter specifies the data U8 (unsigned char) C3 (three

    // channels)

    // 2nd number of columns in the storage

    // 3rd number of rows in the storage

    // 4th underlying storage type - currently only Buffer2D supported

    auto data =

        visioncpp::terminal(in_rgb.get());

    auto data_out =

        visioncpp::terminal(out_hsv.get());



    // unsigned char -> float RGB storage conversion

    auto node = visioncpp::point_operation(data);

    // float RGB to float HSV conversion

    auto node2 = visioncpp::point_operation(node);

    // helper node that allows display of HSV

    // for unsigned char: V <- 255*V, S <- 255*S, H <- H/2 ( to fit in range of 0..255 )

    auto node3 = visioncpp::point_operation(node2);



    // assign operation that writes output of the pipe to output terminal node

    auto pipe = visioncpp::assign(data_out, node3);

    // execute the pipeline

    // 1st template parameter defines if VisionCpp back-end fuses the expression

    // 2nd & 3rd shared memory sizes ( column, row )

    // 4th & 5th local work group size ( column , row )

    visioncpp::execute(pipe, dev);

  }



  printf("RGB: %u %u %u \nHSV: %u %u %u \n", in_rgb.get()[0], in_rgb.get()[1],

         in_rgb.get()[2], out_hsv.get()[0], out_hsv.get()[1], out_hsv.get()[2]);



~~~~~~~~~~~~~~~



## Requirements

To successfully compile VisionCpp tests, you will need:

* ComputeCpp (https://codeplay.com/products/computesuite/computecpp)

* OpenCV 3.2 (https://github.com/opencv/opencv) - used for camera access, window display and as a testing reference.

* GTest (https://github.com/google/googletest) - testing framework.

* OpenCL 1.2



## Build

Assuming you are in the root of a git repo:

~~~~~~~~~~~~~~~{.sh}

mkdir build

cd  build

cmake .. -DComputeCpp_DIR={PATH_TO_COMPUTECPP_ROOT} -DCMAKE_CXX_COMPILER={FAVORITE_CXX_COMPILER}

make -j8

make test

~~~~~~~~~~~~~~~



The output binaries will be catalogued in bin folder.

~~~~~~~~~~~~~~~{.sh}

| - build

  | - bin

    | - example

    | - test

~~~~~~~~~~~~~~~



## Examples

There is a set of example code in the /example/ folder of the repository. Most of the examples are performing image operations from the camera input.



## Documentation

Online documentation can be found [here](https://codeplaysoftware.github.io/visioncpp/).



The documentation is created using [Doxygen](http://www.stack.nl/~dimitri/doxygen/).

~~~~~~~~~~~~~~~{.sh}

make doc

~~~~~~~~~~~~~~~



The documentation will be created in html folder in build directory.

~~~~~~~~~~~~~~~{.sh}

| - build

  | - doc

~~~~~~~~~~~~~~~



## Contributing

Contributors always welcome! See CONTRIBUTING.md for details.



The list of contributors.



## Resources

* [Kernel composition in SYCL](http://opus.bath.ac.uk/49695/)

* [VisionCPP: A SYCL-based Computer Vision Framework](http://dl.acm.org/citation.cfm?doid=2909437.2909444)

* [Wiki](https://github.com/codeplaysoftware/visioncpp/wiki)



## License

The Apache License, Version 2.0 License. See [LICENSE](https://github.com/codeplaysoftware/visioncpp/blob/master/LICENSE) for more.



## Known Issues

* The Tuple class works only with clang++.