作者：CSDN @ _养乐多_

本文将介绍如何实现一个可视化图片和标签信息的查看器，代码使用python实现。点击下一张和上一张可以切换图片。

一、脚本界面

界面如下图所示，

二、完整代码

使用代码时，需要修改 class_id_to_name 还有 YOLO 格式的图片（images）文件夹路径和标签（labels）文件夹路径。

from PIL import Image, ImageDraw, ImageFont, ImageTk
import tkinter as tk
from tkinter import ttk
import os

# 创建类别 ID 到中文名称的映射
class_id_to_name = {
    0: "飞机",
    1: "船只",
    2: "储油罐",
    3: "棒球场",
    4: "网球场",
    5: "篮球场",
    6: "跑道场地",
    7: "港口",
    8: "桥梁",
    9: "车辆"
}

def get_image_size(image_path):
    # 打开图片文件
    with Image.open(image_path) as img:
        # 获取图片的宽度和高度
        width, height = img.size
        return width, height

def read_yolo_labels(label_file, img_width, img_height):
    with open(label_file, 'r') as file:
        lines = file.readlines()

    boxes = []
    for line in lines:
        parts = line.strip().split()
        class_id = int(parts[0])
        x_center = float(parts[1])
        y_center = float(parts[2])
        width = float(parts[3])
        height = float(parts[4])

        # 将 YOLO 格式转换为像素坐标
        x_center_px = int(x_center * img_width)
        y_center_px = int(y_center * img_height)
        width_px = int(width * img_width)
        height_px = int(height * img_height)

        # 计算矩形框的左上角和右下角点
        x1 = int(x_center_px - width_px / 2)
        y1 = int(y_center_px - height_px / 2)
        x2 = int(x_center_px + width_px / 2)
        y2 = int(y_center_px + height_px / 2)

        boxes.append((x1, y1, x2, y2, class_id))

    return boxes

def draw_boxes_on_image(image_path, boxes):
    # 使用 PIL 加载图片
    img = Image.open(image_path)
    draw = ImageDraw.Draw(img)

    # 定义颜色和线宽
    box_color = "yellow"  # 选择一个亮色
    line_width = 5  # 设置较粗的线宽

    # 使用支持中文字符的系统字体
    try:
        # 尝试使用支持中文的常见系统字体
        font = ImageFont.truetype("msyh.ttc", size=24)  # 微软雅黑
    except IOError:
        # 回退到默认字体
        font = ImageFont.load_default()

    for (x1, y1, x2, y2, class_id) in boxes:
        # 绘制矩形框
        draw.rectangle([x1, y1, x2, y2], outline=box_color, width=line_width)
        # 从 class_id 获取类别名称
        class_name = class_id_to_name.get(class_id, "未知")
        text = class_name
        text_width, text_height = 50, 40  # 设定文本框的宽度和高度
        text_x = x1
        text_y = y1 - text_height - 5
        # 绘制带背景矩形的文本
        draw.rectangle([text_x, text_y, text_x + text_width, text_y + text_height], fill=box_color)
        draw.text((text_x, text_y), text, fill="black", font=font)

    return img

def display_image_with_boxes(image_file, label_file):
    # 获取图片尺寸
    img_width, img_height = get_image_size(image_file)

    # 读取 YOLO 标签
    boxes = read_yolo_labels(label_file, img_width, img_height)

    # 在图片上绘制矩形框
    img_with_boxes = draw_boxes_on_image(image_file, boxes)

    return img_with_boxes

class ImageViewer:
    def __init__(self, root, image_files, label_files):
        self.root = root
        self.image_files = image_files
        self.label_files = label_files
        self.current_index = 0

        # 设置固定的查看器大小
        self.viewer_width = 800
        self.viewer_height = 600

        # 初始化界面
        self.init_ui()

    def init_ui(self):
        self.canvas = tk.Canvas(self.root, width=self.viewer_width, height=self.viewer_height)
        self.canvas.pack()

        self.prev_button = ttk.Button(self.root, text="上一张", command=self.prev_image)
        self.prev_button.pack(side=tk.LEFT)

        self.next_button = ttk.Button(self.root, text="下一张", command=self.next_image)
        self.next_button.pack(side=tk.RIGHT)

        self.update_image()

    def update_image(self):
        image_file = self.image_files[self.current_index]
        label_file = self.label_files[self.current_index]
        img_with_boxes = display_image_with_boxes(image_file, label_file)

        # 将图片转换为 Tkinter 可用格式
        img_with_boxes = img_with_boxes.convert("RGB")
        img_tk = ImageTk.PhotoImage(img_with_boxes)

        # 计算缩放比例
        img_width, img_height = img_with_boxes.size
        scale = min(self.viewer_width / img_width, self.viewer_height / img_height)
        new_width = int(img_width * scale)
        new_height = int(img_height * scale)

        # 缩放图片
        img_resized = img_with_boxes.resize((new_width, new_height), Image.Resampling.LANCZOS)
        img_tk_resized = ImageTk.PhotoImage(img_resized)

        # 清除画布上的内容
        self.canvas.delete("all")

        # 在画布上显示图片
        self.canvas.create_image(self.viewer_width / 2, self.viewer_height / 2, image=img_tk_resized)

        # 保持对图像的引用
        self.canvas.image = img_tk_resized

    def prev_image(self):
        if self.current_index > 0:
            self.current_index -= 1
            self.update_image()

    def next_image(self):
        if self.current_index < len(self.image_files) - 1:
            self.current_index += 1
            self.update_image()

if __name__ == "__main__":
    # 图片和标签文件的路径
    image_folder = 'E:\\DataSet\\positive'
    label_folder = 'E:\\DataSet\\yolo_labels'

    # 获取所有图片和标签文件
    image_files = sorted([os.path.join(image_folder, f) for f in os.listdir(image_folder) if f.endswith('.jpg')])
    label_files = sorted([os.path.join(label_folder, f) for f in os.listdir(label_folder) if f.endswith('.txt')])

    # 创建 Tkinter 窗口
    root = tk.Tk()
    root.title("图片标注查看器")

    # 启动图像查看器
    viewer = ImageViewer(root, image_files, label_files)
    root.mainloop()