# INSTALL

* OS : `Window 10`
* GPU : `RTX-2080 Ti`
* Visual Studio : `Community 2019`
* Opencv : `3.4.10`
* CUDA : `10.1`
* cudnn : `7.65`

GPU 버전에 맞게 CUDA와 cudnn을 설치를 하시면 됩니다.

* CUDA : <https://developer.nvidia.com/cuda-toolkit-archive>
* cudnn : <https://developer.nvidia.com/rdp/form/cudnn-download-survey>
* 환경변수 등록 !!

## CUDA version 확인

```shell
nvcc --version
```

## cudnn version 확인

`C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\include\cudnn.h`

```c
#define CUDNN_MAJOR 7
#define CUDNN_MINOR 6
#define CUDNN_PATCHLEVEL 5
```

version : `7.65`

***

## Visual Studio 설치

* <https://visualstudio.microsoft.com/ko/>

1. Visual Studio 다운로드 -> Community 2019
2. C++를 사용한 데스크톱 개발 Check
3. 설치

## Opencv 설치

* <https://opencv.org/releases/>
* `3.4.10`
* 환경변수 등록 !!

## Darknet 설치

```shell
git clone https://github.com/AlexeyAB/darknet.git
```

## DarkNet 실행하기

1. `darknet/build/darknet/darknet.sln` 접속
2. `프로젝트` -> `속성`

* 2.0 `구성 : Releases / 플랫폼 : x64`
* 2.1 `C/C++` -> `일반` -> `추가 포함 디렉터리` -> `opencv 경로 추가 (~\opencv\build\include)`
* 2.2 `전처리기` -> `전처리기 정의` -> `OPENCV 추가`
* 2.3 `링커` -> `일반` -> `추가 라이브러리 디렉터리` -> `opencv 경로 추가 (~\opencv\build\x64\vc14\lib)`
* 2.4 `구성 : Debug / 플랫폼 : x64`에도 똑같이 적용

3. `솔루션 탐색기` -> `darknet 우클릭` -> `빌드 종속성` -> `사용자 지정 빌드` -> `CUDA Check`
4. `~\opencv\build\x64\vc14\bin` -> `opencv_ffmpeg3410_64.dll`, `opencv_world3410.dll` 복사 -> `~\darknet\build\darknet\x64` 디렉토리에 붙여넣기
5. `ctrl + F5`
6. `빌드 성공 ^^`

## 문제점 해결

* CUDA 10.1인 경우 처음에 빌드 종속성에서 찾을 수 없습니다.

[여기](https://bigcode.tistory.com/47)를 참조해 따라하고 빌드 종속성을 `CUDA 10.1`로 변경하시면 됩니다.

## 예제 실행

* yolov3 weights : <https://pjreddie.com/media/files/yolov3.weights>
* tiny yolov3 weights : <https://pjreddie.com/media/files/yolov3-tiny.weights>

```shell
cd ~/darknet/build/darknet/x64
```

```shell
darknet.exe detector test data/coco.data cfg/yolov3.cfg yolov3.weights -i 0 dog.jpg
```


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://opensource-book.gitbook.io/darknet-book-2/install.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.