使用yolo格式标注将目标框中的图片裁剪保存

最新推荐文章于 2024-07-04 19:26:49 发布

Sean X

最新推荐文章于 2024-07-04 19:26:49 发布

阅读量1.2k

点赞数 3

文章标签： YOLO opencv 人工智能计算机视觉目标检测

本文链接：https://blog.csdn.net/m0_56182552/article/details/132760251

版权

本文介绍了一个Python程序，用于根据YOLO格式的txt文件对图片进行目标框截取，将指定区域的图片保存到新的目录中。它遍历图片和txt文件，读取坐标信息，然后使用OpenCV进行坐标转换和裁剪操作。

摘要由CSDN通过智能技术生成

根据yolo格式标注的txt文件，将目标框中的图片截取保存

Python代码：

import os
import cv2

def txt_cut_image(image_path,txt_path, new_image_path):
    # 遍历所有图片
    for file in os.listdir(image_path):
            image_file_path = os.path.join(image_path, file)

            # 构造对应的txt文件路径
            txt_file = os.path.splitext(file)[0] + ".txt"
            label_file_path = os.path.join(txt_path, txt_file)

            # 读取标注文件和原始图片
            with open(label_file_path, 'r') as f:
                lines = f.readlines()
            image = cv2.imread(image_file_path)

            # 循环处理每个标注框
            i = 0
            for line in lines:
                i = i + 1
                class_id, x_center, y_center, width, height = map(float, line.strip().split())

                # 将YOLOv5格式的坐标转换为常规坐标
                left = int((x_center - width / 2) * image.shape[1])
                top = int((y_center - height / 2) * image.shape[0])
                right = int((x_center + width / 2) * image.shape[1])
                bottom = int((y_center + height / 2) * image.shape[0])

                # 截取标注框内的内容并保存为新图片
                cut_image = image[top:bottom, left:right]
                new_cut_image_path = os.path.join(new_image_path,file)
                print(new_cut_image_path)
                cv2.imwrite(new_cut_image_path, cut_image)

if __name__ == "__main__":
    image_path = "./data/images"
    txt_path = "./data/labels"
    new_image_path = "./data/new_images"
    txt_cut_image(image_path,txt_path, new_image_path)