一、环境配置

• Ubuntu 22.04
• Torch 1.10
• CUDA 11.3
• python 3.9

      环境配置参考下面链接（建议Linux系统）yolov5_obb/install.md at master · hukaixuan19970627/yolov5_obb (github.com)https://github.com/hukaixuan19970627/yolov5_obb/blob/master/docs/install.md

•  创建虚拟环境

conda create -n Py39_Torch1.10_cu11.3 python=3.9 -y 
source activate Py39_Torch1.10_cu11.3

•  查看CUDA版本（确保CUDA runtime api version ≤ CUDA driver version）

nvcc -V
nvidia-smi

•  安装PyTorch和torchvision

pip3 install torch==1.10.1+cu113 torchvision==0.11.2+cu113 torchaudio==0.10.1+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html
nvcc -V
python
>>> import torch
>>> torch.version.cuda
>>> exit()

•  克隆yolov7-obb项目（也可直接在GitHub上下载）

git clone https://github.com/Egrt/yolov7-obb.git

•  配置yolov7-obb所需环境

cd yolov7-obb
pip install -r requirements.txt
cd utils/nms_rotated
python setup.py develop #安装非极大值抑制库

运行 python setup.py develop时报错

可以看出时gcc版本问题，检测已有版本发现gcc 12 > gcc 11

故安装低版本的gcc，并创建软链接（此处要对应改为自己的位置）

sudo apt-get install gcc-10
sudo apt-get install g++-10sudo ln -s /usr/bin/gcc-10 /usr/local/cuda-11.6/bin/gcc #创建软链接

再次运行 python setup.py develop

至此，环境配置完成！！！

二、数据集制作

• 下载官方源码

https://github.com/cgvict/roLabelImg

•  进入下载好的roLabellmg-master文件夹内 ，在终端打开

pyrcc5 -oresources.py resources.qrc
python roLabelImg.py

•  快捷键

•  标注的XML格式

hsrc100000001/Users/haoyou/Library/Mobile Documents/com~apple~CloudDocs/OneDrive/hsrc/100000001.bmpUnknown116675330

•  数据集格式转换（把旋转框 cx,cy,w,h,angle，转换成四点坐标x1,y1,x2,y2,x3,y3,x4,y4）

import os
import xml.etree.ElementTree as ET
import mathdef edit_xml(xml_file, dotaxml_file):"""修改xml文件:param xml_file:xml文件的路径:return:"""tree = ET.parse(xml_file)objs = tree.findall('object')for ix, obj in enumerate(objs):x0 = ET.Element("x0")  # 创建节点y0 = ET.Element("y0")x1 = ET.Element("x1")y1 = ET.Element("y1")x2 = ET.Element("x2")y2 = ET.Element("y2")x3 = ET.Element("x3")y3 = ET.Element("y3")# obj_type = obj.find('bndbox')# type = obj_type.text# print(xml_file)if (obj.find('robndbox') == None):obj_bnd = obj.find('bndbox')obj_xmin = obj_bnd.find('xmin')obj_ymin = obj_bnd.find('ymin')obj_xmax = obj_bnd.find('xmax')obj_ymax = obj_bnd.find('ymax')xmin = float(obj_xmin.text)ymin = float(obj_ymin.text)xmax = float(obj_xmax.text)ymax = float(obj_ymax.text)obj_bnd.remove(obj_xmin)  # 删除节点obj_bnd.remove(obj_ymin)obj_bnd.remove(obj_xmax)obj_bnd.remove(obj_ymax)x0.text = str(xmin)y0.text = str(ymax)x1.text = str(xmax)y1.text = str(ymax)x2.text = str(xmax)y2.text = str(ymin)x3.text = str(xmin)y3.text = str(ymin)else:obj_bnd = obj.find('robndbox')obj_bnd.tag = 'bndbox'  # 修改节点名obj_cx = obj_bnd.find('cx')obj_cy = obj_bnd.find('cy')obj_w = obj_bnd.find('w')obj_h = obj_bnd.find('h')obj_angle = obj_bnd.find('angle')cx = float(obj_cx.text)cy = float(obj_cy.text)w = float(obj_w.text)h = float(obj_h.text)angle = float(obj_angle.text)obj_bnd.remove(obj_cx)  # 删除节点obj_bnd.remove(obj_cy)obj_bnd.remove(obj_w)obj_bnd.remove(obj_h)obj_bnd.remove(obj_angle)x0.text, y0.text = rotatePoint(cx, cy, cx - w / 2, cy - h / 2, -angle)x1.text, y1.text = rotatePoint(cx, cy, cx + w / 2, cy - h / 2, -angle)x2.text, y2.text = rotatePoint(cx, cy, cx + w / 2, cy + h / 2, -angle)x3.text, y3.text = rotatePoint(cx, cy, cx - w / 2, cy + h / 2, -angle)# obj.remove(obj_type)  # 删除节点obj_bnd.append(x0)  # 新增节点obj_bnd.append(y0)obj_bnd.append(x1)obj_bnd.append(y1)obj_bnd.append(x2)obj_bnd.append(y2)obj_bnd.append(x3)obj_bnd.append(y3)tree.write(dotaxml_file, method='xml', encoding='utf-8')  # 更新xml文件# 转换成四点坐标
def rotatePoint(xc, yc, xp, yp, theta):xoff = xp - xc;yoff = yp - yc;cosTheta = math.cos(theta)sinTheta = math.sin(theta)pResx = cosTheta * xoff + sinTheta * yoffpResy = - sinTheta * xoff + cosTheta * yoffreturn str(int(xc + pResx)), str(int(yc + pResy))def totxt(xml_path, out_path):# 想要生成的txt文件保存的路径，这里可以自己修改files = os.listdir(xml_path)for file in files:tree = ET.parse(xml_path + os.sep + file)root = tree.getroot()name = file.strip('.xml')output = out_path + name + '.txt'file = open(output, 'w')objs = tree.findall('object')for obj in objs:cls = obj.find('name').textbox = obj.find('bndbox')x0 = int(float(box.find('x0').text))y0 = int(float(box.find('y0').text))x1 = int(float(box.find('x1').text))y1 = int(float(box.find('y1').text))x2 = int(float(box.find('x2').text))y2 = int(float(box.find('y2').text))x3 = int(float(box.find('x3').text))y3 = int(float(box.find('y3').text))file.write("{} {} {} {} {} {} {} {} {} 0\n".format(x0, y0, x1, y1, x2, y2, x3, y3, cls))file.close()print(output)if __name__ == '__main__':# -----**** 第一步：把xml文件统一转换成旋转框的xml文件 ****-----roxml_path = "./datasets/Annotations"  # 目录下保存的是需要转换的xml文件dotaxml_path = './datasets/dotaxml'filelist = os.listdir(roxml_path)for file in filelist:edit_xml(os.path.join(roxml_path, file), os.path.join(dotaxml_path, file))

三、训练自己的数据集

      本文参考yolov7-obb项目如下Egrt/yolov7-obb: 在YOLOv7的基础上使用KLD损失修改为旋转目标检测yolov7-obb (github.com)https://github.com/Egrt/yolov7-obb

• 数据集格式 （其中Annotations为上面格式转换过的xml文件）

VOCdevkit/VOC2007├── Annotations├── 0001.xml├── 0002.xml..├── ImageSets├── Main├── JPEGImages├── 0001.xml├── 0002.xml..

•  运行voc_annotation.py文件，生成2007_train.txt和2007_val.txt文件

•  开始训练