目录
YOLOv8 是 YOLO 系列最新的目标检测模型，具有高效性和高准确度。本文将详细介绍如何在本地环境中配置 YOLOv8，包括环境搭建、依赖安装、模型下载和基本使用示例。
本系列其他文章
【YOLOv8系列】（一）YOLOv8介绍：实时目标检测的最新突破-CSDN博客
一. 准备环境
1.Anaconda下载
Anaconda 是一个开源的 Python 和 R 语言的发行版，致力于简化数据科学、机器学习、人工智能和大数据的包管理和部署。它包含了大量流行的数据科学包，并且通过 Conda 包管理器提供了便捷的包安装、更新和管理方式。
选择适合自己系统的Anaconda软件下载，然后按照步骤安装在除C盘以外的磁盘。
注意：在这一步骤需要勾选以上选项，其他步骤默认点击next安装即可
2.创建yolov8虚拟环境
在电脑左下角搜索Anaconda，点击打开Anaconda Prompt
创建新的虚拟环境yolov8
conda create -n yolov8 python=3.8
激活yolov8环境
conda activate yolov8
3.pytorch安装
Win+R输入cmd命令弹出对话框后。输入命令，nvidia-smi自己电脑是否有GPU，如果有，CUDA版本是多少。如果没有则安装CPU版本的pytoch，跳过后续CUDA以及CUDNN步骤。
nvidia-smi
进入PyTorch官网，查看对应自己电脑情况的安装命令PyTorch
https://pytorch.org/
4.CUDA下载
打开CUDA Toolkit Archive | NVIDIA Developer官网
CUDA Toolkit Archive | NVIDIA Developerhttps://developer.nvidia.com/cuda-toolkit-archive 选择自己电脑版本的CUDA，例如CUDA Toolkit 11.6.1
依次选择系统等信息，然后点击下载到电脑。默认安装即可，记录下安装路径。
安装完成后，【右击此电脑】—>【属性】—>(界面右边)【高级系统设置】—>【环境变量】—>点击【系统变量】里的path进去查看环境变量。需要有一下环境变量，若没有安装默认安装路径自行添加。
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6\libnvvp
C:\Program Files\NVIDIA Corporation\Nsight Compute 2022.1.1\
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6\bin
5.CUDNN下载
将下载解压后将所有文件拷贝到CUDA目录里，例如CUDA默认路径为C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6
二、yolov8模型下载
1.clone模型
可以直接下载zip压缩文件，或者安装Git输入以下命令克隆ultralytics项目
git clone https://github.com/ultralytics/ultralytics.git下载完了之后目录如下：
很多小伙伴说目前GitHub只有v11版本了，这里博主提供v8版本资源供大家下载
2.pycharm配置
①解释器配置
将下载下来的ultralytics文件夹使用pycharm打开，进入文件 —> 设置界面，找到解释器选项。选择你希望使用的 Anaconda 环境作为解释器，并进行配置。
点击添加本地解释器—>现有—>yolov8的conda虚拟环境下的python.exe
②终端配置
打开Anaconda Prompt所在文件，右键属性查看目标位置，复制cmd.exe以及后面的内容。
同样地，文件—>设置—>工具—>终端，将以上复制的路径填入shell路径。
3.安装必要库
在pycharm终端输入pip安装命令等待安装完成。
pip install ultralytics -i https://pypi.tuna.tsinghua.edu.cn/simple
pip install yolo -i https://pypi.tuna.tsinghua.edu.cn/simple
4.下载训练模型
在GitHub项目仓，README往下翻可以看见预测、分类、分割等模型，点击即可下载。https://github.com/ultralytics/ultralyticshttps://github.com/ultralytics/ultralytics
或者如果网络不太好，可以使用脚本下载或者直接下载我上传的资源。
首先需要新建一个model.txt，包含需要下载模型的地址。
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8n-pose.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8s-pose.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8m-pose.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8l-pose.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8x-pose.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8n-seg.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8s-seg.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8m-seg.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8l-seg.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8x-seg.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8n-cls.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8s-cls.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8m-cls.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8l-cls.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8x-cls.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8n.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8s.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8m.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8l.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8x.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8n-obb.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8s-obb.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8m-obb.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8l-obb.pt
https://github.com/ultralytics/assets/releases/download/v8.2.0/yolov8x-obb.pt然后运行以下python脚本进行下载。
#!usr/bin/env python
# encoding:utf-8
from __future__ import division
import os
import time
import logging
import subprocess
import urllib
import requests
import torch
import random
from pathlib import Path
def autoDownload(filePath, downloadDir):
if not os.path.exists(downloadDir):
os.makedirs(downloadDir)
while True:
with open(filePath) as f:
lists = [one.strip() for one in f.readlines() if one.strip()]
count = 0
for downloadUrl in lists:
print("模型下载地址: ", downloadUrl)
try:
tmpFile = downloadDir + downloadUrl.split("/")[-1].strip()
print("下载文件路径: ", tmpFile)
if not os.path.exists(tmpFile):
torch.hub.download_url_to_file(downloadUrl, str(tmpFile))
else:
print("当前训练模型文件已经下载完成!")
count += 1
except Exception as e:
print("Exception: ", e)
time.sleep(random.randint(1, 5))
print("下载数量为: ", count)
if count == len(lists):
break
if __name__ == "__main__":
autoDownload(filePath="model.txt", downloadDir="model/")等待.pt模型下载完毕即可
三、 环境验证
在终端输入以下命令进行预测，打开runs/detect/predict查看预测结果
yolo predict model=./model/yolov8n.pt source='https://ultralytics.com/images/bus.jpg'
可以看到yolov8的效果是非常棒的，除了person以及bus，甚至连stop sign都检测出来了。
四、总结
通过以上步骤，你应该已经成功配置好YOLOv8的环境。接下来你可以进行模型训练和推理，期待你在目标检测任务中取得好成绩！如果有任何问题，欢迎在评论区讨论。
如果以上内容对您有帮助，可以三连打赏订阅本专栏哦， 谢谢~
扫一扫
6万+
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
