利用旋转平移缩放裁剪，高斯椒盐，均匀，高斯模糊方法扩充数据集并生成xml，txt 文件代码

在图像处理领域中，使用这些技术扩充数据集通常是在计算机视觉任务如物体识别、行人检测等场景。这里提供一个基于Python和OpenCV的基本框架，展示了如何对图片进行操作后保存成XML（例如PASCAL VOC格式，用于标注）和TXT文件。请注意，这只是一个简化版本，实际应用可能需要更复杂的图像处理库如`imgaug`。

import cv2
import numpy as np
from lxml import etree
import os

# 假设你有一个基础图片目录base_dir
base_dir = 'base_images'
output_xml_dir = 'output_xml'
output_txt_dir = 'output_txt'

def rotate_scale_crop(image, angle=(-5, 5), scale=(0.9, 1.1)):
    # 随机旋转、缩放和平移
    h, w = image.shape[:2]
    center = (w // 2, h // 2)
    rotation_matrix = cv2.getRotationMatrix2D(center, random.uniform(angle[0], angle[1]), scale[0])
    image = cv2.warpAffine(image, rotation_matrix, (w, h))
    top_left = (np.random.randint(0, int((1-scale[1]) * w)), np.random.randint(0, int((1-scale[1]) * h)))
    bottom_right = (np.random.randint(int(scale[1] * w), w), np.random.randint(int(scale[1] * h), h))
    return image[top_left[1]:bottom_right[1], top_left[0]:bottom_right[0]]

def add_gaussian_noise(image, sigma=1):
    noise = np.random.normal(loc=0, scale=sigma, size=image.shape)
    noisy_image = image + noise
    return noisy_image.astype(np.uint8)

def augment_image(image, save_path):
    # 先旋转和裁剪
    img_augmented = rotate_scale_crop(image)
    
    # 然后添加高斯噪声
    img_augmented = add_gaussian_noise(img_augmented)

    # ... 这里也可以加入其他滤波效果（如高斯模糊）

    # 保存图片到目标路径
    cv2.imwrite(save_path, img_augmented)

    # 生成XML标注（PASCAL VOC格式）
    root = etree.Element("annotation")
    folder = etree.SubElement(root, "folder")
    folder.text = 'test'
    filename = etree.SubElement(root, "filename")
    filename.text = os.path.basename(save_path)
    size = etree.SubElement(root, "size")
    width = etree.SubElement(size, "width")
    width.text = str(img_augmented.shape[1])
    height = etree.SubElement(size, "height")
    height.text = str(img_augmented.shape[0])
    depth = etree.SubElement(size, "depth")
    depth.text = "3"

    # 添加随机坐标框信息
    x, y, w, h = np.random.randint(0, high=img_augmented.shape[1], size=4)  # 随机边界框
    obj = etree.SubElement(root, "object")
    name = etree.SubElement(obj, "name")
    name.text = 'example_object'  # 你需要替换为你想要的对象名称
    bndbox = etree.SubElement(obj, "bndbox")
    xmin = etree.SubElement(bndbox, "xmin")
    xmin.text = str(x)
    ymin = etree.SubElement(bndbox, "ymin")
    ymin.text = str(y)
    xmax = etree.SubElement(bndbox, "xmax")
    xmax.text = str(x+w)
    ymax = etree.SubElement(bndbox, "ymax")
    ymax.text = str(y+h)

    xml_file = os.path.splitext(save_path)[0] + ".xml"
    xml_path = os.path.join(output_xml_dir, xml_file)
    with open(xml_path, 'wb') as f:
        f.write(etree.tostring(root, pretty_print=True))

    # 生成TXT文件
    txt_file = os.path.splitext(save_path)[0] + ".txt"
    txt_path = os.path.join(output_txt_dir, txt_file)
    with open(txt_path, 'w', encoding='utf-8') as f:
        f.write(f"{os.path.basename(xml_file)}\n")

# 遍历基础图片目录，进行数据增强并保存
for img_path in os.listdir(base_dir):
    img = cv2.imread(os.path.join(base_dir, img_path))
    augment_image(img, os.path.join(output_xml_dir, img_path.split('.')[0] + '_augmented.jpg'))