---

前言

YOLOv8是Ultralytics的最新版本YOLO。作为最先进的 SOTA 模型，YOLOv8 建立在以前版本成功的基础上，引入了新功能和改进，以增强性能、灵活性和效率。YOLOv8 支持全方位的视觉 AI 任务，包括检测、分割、姿势估计、跟踪和分类。这种多功能性使用户能够在不同的应用程序和域中利用YOLOv8的功能。模型中只需要设定不同的训练模型，就可以得到不同的检测结果

1、环境安装

2.1安装torch相关库

官网地址：https://pytorch.org/get-started/locally

安装命令：（根据自己电脑选型）

pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

2.2 获取yolov8最新版本，并安装依赖

github地址： https://github.com/ultralytics/ultralytics

通过克隆仓库到本地来获得yolov8最新版本。

安装依赖环境

pip install ultralytics

3. 如何使用模型用于各种CV任务

安装好yolov8需要的运行环境之后，就可以直接使用模型了，目前，该模型可以直接用于检测、分割、姿势估计、跟踪和分类这5类检测任务。
支持两种运行方式：

方式一：命令行形式

yolo TASK MODE ARGS

举例： yolo detect predict model=yolov8n.pt source='test.jpg'
表示对图片test.jpg进行目标检测

参数说明：

TASK (optional) 表示模型任务类型，可以从列表 [detect, segment, classify, pose]选一个.
分别代表检测、分割、分类、姿态检测任务 MODE (required) 表示模型需要执行的操作，可以从列表 [train, val,
predict, export, track, benchmark]选一个，分别代表训练、验证、预测、模型转换、追踪、基准模型评估 ARGS
(optional) 表示其他的一些参数设置，后面详细说明

方式二：python代码形式示例

from ultralytics import YOLO

# 加载预训练模型，
model = YOLO('yolov8n.pt',task='detect')
# yolov8n.pt表示预训练模型
# task代表需要执行的任务：detect, segment, classify, pose
# 使用模型进行指定任务
results = model(ARGS)

3.1 目标检测任务实现

代码

from ultralytics import YOLO
import cv2
# 加载预训练模型
model = YOLO("yolov8n.pt", task='detect') 
# model = YOLO("yolov8n.pt") task参数也可以不填写，它会根据模型去识别相应任务类别
# 检测图片
results = model("./ultralytics/assets/bus.jpg")
res = results[0].plot()
cv2.imshow("YOLOv8 Inference", res)
cv2.waitKey(0)

运行结果

检测视频代码

import cv2
from ultralytics import YOLO

# Load the YOLOv8 model
model = YOLO('yolov8n.pt')
print('111')
# Open the video file
video_path = "1.mp4"
cap = cv2.VideoCapture(video_path)

# Loop through the video frames
while cap.isOpened():
    # Read a frame from the video
    success, frame = cap.read()

    if success:
        # Run YOLOv8 inference on the frame
        results = model(frame)

        # Visualize the results on the frame
        annotated_frame = results[0].plot()

        # Display the annotated frame
        cv2.imshow("YOLOv8 Inference", annotated_frame)

        # Break the loop if 'q' is pressed
        if cv2.waitKey(1) & 0xFF == ord("q"):
            break
    else:
        # Break the loop if the end of the video is reached
        break

# Release the video capture object and close the display window
cap.release()
cv2.destroyAllWindows()

3.2 分割任务实现

代码

from ultralytics import YOLO
import cv2
# Load a model
model = YOLO('yolov8n-seg.pt')  
# Predict with the model
results = model('./ultralytics/assets/bus.jpg')  # predict on an image
res = results[0].plot(boxes=False) #boxes=False表示不展示预测框，True表示同时展示预测框
# Display the annotated frame
cv2.imshow("YOLOv8 Inference", res)
cv2.waitKey(0)

运行效果

分割视频代码

import cv2
from ultralytics import YOLO

# Load the YOLOv8 model
model = YOLO('yolov8n-seg.pt', task='segment')
# Open the video file
video_path = "1.mp4"
cap = cv2.VideoCapture(video_path)

# Loop through the video frames
while cap.isOpened():
    # Read a frame from the video
    success, frame = cap.read()

    if success:
        # Run YOLOv8 inference on the frame
        results = model(frame)

        # Visualize the results on the frame
        annotated_frame = results[0].plot()
        # annotated_frame = results[0].plot(boxes=False)不显示预测框

        # Display the annotated frame
        cv2.imshow("YOLOv8 Inference", annotated_frame)

        # Break the loop if 'q' is pressed
        if cv2.waitKey(1) & 0xFF == ord("q"):
            break
    else:
        # Break the loop if the end of the video is reached
        break

# Release the video capture object and close the display window
cap.release()
cv2.destroyAllWindows()

3.3 追踪任务

代码如下：【与目标检测不同的是，每个物体有一个ID。】

代码

from ultralytics import YOLO

# Load a model
model = YOLO('yolov8n.pt',task='detect')  
# model = YOLO('yolov8n-seg.pt')  

# Track with the model
results = model.track(source="1.mp4", show=True)

3.4 姿态检测任务

代码

from ultralytics import YOLO
import cv2
# Load a model
model = YOLO('yolov8n-pose.pt')  # load a pretrained model (recommended for training)
results = model('./ultralytics/assets/bus.jpg')
res = results[0].plot()
cv2.imshow("YOLOv8 Inference", res)
cv2.waitKey(0)

运行结果

姿态检测视频代码

import cv2
from ultralytics import YOLO

# Load the YOLOv8 model
model = YOLO('yolov8n-pose.pt', task='pose')
# Open the video file
video_path = "1.mp4"
cap = cv2.VideoCapture(video_path)

# Loop through the video frames
while cap.isOpened():
    # Read a frame from the video
    success, frame = cap.read()

    if success:
        # Run YOLOv8 inference on the frame
        results = model(frame)

        # Visualize the results on the frame
        annotated_frame = results[0].plot()

        # Display the annotated frame
        cv2.imshow("YOLOv8 Inference", annotated_frame)

        # Break the loop if 'q' is pressed
        if cv2.waitKey(1) & 0xFF == ord("q"):
            break
    else:
        # Break the loop if the end of the video is reached
        break

# Release the video capture object and close the display window
cap.release()
cv2.destroyAllWindows()