在YOLOv8中实现图像上传与关键点检测结果可视化

# 确保已安装ultralytics库
# !pip install ultralytics

from ultralytics import YOLO
import matplotlib.pyplot as plt
import matplotlib.image as mpimg
import os
from google.colab import files # 仅适用于Google Colab环境

# 加载预训练的关键点检测模型
# 您可以使用YOLOv8提供的预训练模型，例如yolov8n-pose.pt
# 或者您自己训练的模型
model = YOLO('yolov8n-pose.pt') 

print("YOLOv8模型加载成功。")

def upload_image():
    """
    在Google Colab环境中上传图片，并返回上传图片的路径。
    """
    uploaded = files.upload()
    for filename in uploaded.keys():
        print(f'用户已上传文件: {filename}')
        return filename # 通常只处理一个文件

    return None

# 调用上传函数
uploaded_image_path = upload_image()

if uploaded_image_path:
    print(f"准备处理图片: {uploaded_image_path}")
else:
    print("未上传图片。")

def run_inference_and_save(image_path, model):
    """
    对上传的图片运行YOLOv8关键点检测推理，并保存结果。
    """
    if not image_path or not os.path.exists(image_path):
        print(f"错误：图片路径无效或文件不存在: {image_path}")
        return None

    print(f"正在对图片 '{image_path}' 进行推理...")

    # 运行推理，并设置 save=True 以保存带标注的图片
    results = model(image_path, save=True, conf=0.25) # conf参数可调整检测置信度阈值

    # YOLOv8通常将结果保存在 runs/pose/predict/ 目录下
    # 如果是第一次运行，目录可能是 predict，第二次是 predict2，以此类推
    # 我们可以通过 results[0].save_dir 获取实际的保存目录
    if results and len(results) > 0:
        save_dir = results[0].save_dir
        # 结果图像的名称通常与输入图像相同
        output_image_name = os.path.basename(image_path)
        processed_image_path = os.path.join(save_dir, output_image_name)
        print(f"处理后的图片已保存至: {processed_image_path}")
        return processed_image_path
    else:
        print("推理未生成任何结果。")
        return None

# 执行推理并获取处理后图片的路径
processed_image_path = run_inference_and_save(uploaded_image_path, model)

def display_image(image_path, title="关键点检测结果"):
    """
    使用matplotlib显示图片。
    """
    if not image_path or not os.path.exists(image_path):
        print(f"错误：无法显示图片，路径无效或文件不存在: {image_path}")
        return

    print(f"正在显示图片: {image_path}")
    img = mpimg.imread(image_path)
    plt.figure(figsize=(10, 8))
    plt.imshow(img)
    plt.title(title)
    plt.axis('off') # 不显示坐标轴
    plt.show()

# 显示处理后的关键点图像
if processed_image_path:
    display_image(processed_image_path, "YOLOv8 关键点检测结果")
else:
    print("无法显示处理后的图片，因为推理未成功或未保存图片。")

import os
from ultralytics import YOLO
import matplotlib.pyplot as plt
import matplotlib.image as mpimg
from google.colab import files # 仅适用于Google Colab

# --- 1. 模型加载 ---
model = YOLO('yolov8n-pose.pt') # 替换为您自己的模型路径

# --- 2. 图像上传 ---
def upload_image_colab():
    uploaded = files.upload()
    if uploaded:
        filename = list(uploaded.keys())[0]
        print(f'用户已上传文件: {filename}')
        return filename
    return None

uploaded_image_path = upload_image_colab()

if uploaded_image_path:
    # --- 3. YOLOv8推理与结果保存 ---
    print(f"正在对图片 '{uploaded_image_path}' 进行推理...")
    results = model(uploaded_image_path, save=True, conf=0.25)

    processed_image_path = None
    if results and len(results) > 0:
        save_dir = results[0].save_dir
        output_image_name = os.path.basename(uploaded_image_path)
        processed_image_path = os.path.join(save_dir, output_image_name)
        print(f"处理后的图片已保存至: {processed_image_path}")
    else:
        print("推理未生成任何结果。")

    # --- 4. 可视化处理后的图像 ---
    if processed_image_path and os.path.exists(processed_image_path):
        print(f"正在显示图片: {processed_image_path}")
        img = mpimg.imread(processed_image_path)
        plt.figure(figsize=(10, 8))
        plt.imshow(img)
        plt.title("YOLOv8 关键点检测结果")
        plt.axis('off')
        plt.show()
    else:
        print("无法显示处理后的图片，文件不存在或路径错误。")
else:
    print("未上传图片，无法进行推理。")