Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/lc-soft/darknetlib
C library for wrapping darknet
https://github.com/lc-soft/darknetlib
c-api c-bindings darknet deep-learning dnn object-detection yolo
Last synced: about 1 month ago
JSON representation
C library for wrapping darknet
- Host: GitHub
- URL: https://github.com/lc-soft/darknetlib
- Owner: lc-soft
- License: mit
- Created: 2018-12-28T14:00:35.000Z (almost 6 years ago)
- Default Branch: master
- Last Pushed: 2019-10-17T14:58:51.000Z (about 5 years ago)
- Last Synced: 2024-04-14T22:44:27.270Z (8 months ago)
- Topics: c-api, c-bindings, darknet, deep-learning, dnn, object-detection, yolo
- Language: C
- Homepage: https://github.com/AlexeyAB/darknet
- Size: 1.08 MB
- Stars: 6
- Watchers: 3
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# DarknetLib
([中文](README.zh-cn.md)/**EN**)
## Tab of contents
- [Introduction](#introduction)
- [Features](#features)
- [Use](#use)
- [Use on Windows](#use-on-windows)
- [Use on Linux](#use-on-linux)
- [Todo](#todo)
- [Legal](#legal)
## Introduction
Darknetlib is a C library for wrapping [darknet](http://pjreddie.com/darknet/),
It mainly provides object detection support for [LC-Finder](https://github.com/lc-soft/LC-Finder), you can find the relevant code in the [src/lib/detector.c](https://github.com/lc-soft/LC-Finder/blob/develop/src/lib/detector.c) file.
## Features
- Provides simple exception handling
- Redefining the interface with a new naming style
- Suitable for compilation into a dynamic library to use
## Use
### Use on Windows
Install with [lcpkg](https://github.com/lc-soft/lcpkg):
```bash
lcpkg install github.com/lc-soft/darknetlib
```
Darknetlib installed in this way is the pure CPU computing version, If you need gpu-accelerated version, please go to [Release](https://github.com/lc-soft/darknetlib/releases) page to download.
If you want to build from source code:
1. Download and install [CUDA](https://developer.nvidia.com/cuda-downloads)
1. Download [cuDNN](https://developer.nvidia.com/cudnn) and unpack files to hte `3rdparty` directory
1. Open `build/darknet.sln` whith Visual Studio 2017
1. Set **Release** and **x64**
1. build **darknet_gpu** and **test** project
1. Download the [yolov3-tiny.weights](https://pjreddie.com/media/files/yolov3-tiny.weights) file to the `test` directory
1. Run the test
### Use on Linux
```bash
git clone https://github.com/lc-soft/darknetlib.git
git submodule init
git submodule update
make
cd test
make
```
## Todo
- Let `darknet_detector_train()` work fine.
- Friendly and standardized log output.
- Remove command line interaction code, like: `getchar()`.
- Remove the strange `system()` call, like: `system("echo ...")`
- Refactor source code of the darknet, make it more suitable for use as a library, not a command line program. (very low priority)
## Legal
Code released under the [MIT License](LICENSE).