使用pybgs库SuBSENSE算法时遇XML文件写入错误求助

问题重现 #

测试pybgs库中的背景分割算法时，SuBSENSE算法无法正常工作，同时抛出文件读写错误。运行代码如下：

import cv2
import pybgs as bgs

video_path = "video.mp4"

# create VideoCapture object for further video processing
captured_video = cv2.VideoCapture(video_path)
# check video capture status
if not captured_video.isOpened:
    print("Unable to open: " + video_path)
    exit(0)

background_sub_method = bgs.SuBSENSE()

while True:
    # read video frames
    ret, frame = captured_video.read()
    # check whether the frames have been grabbed
    if not ret:
        break

    # pass the frame to the background subtractor
    foreground_mask = background_sub_method.apply(frame)
    # obtain the background without foreground mask
    img_bg_model = background_sub_method.getBackgroundModel()

    cv2.imshow("Initial Frame", frame)
    cv2.imshow("FG Mask", foreground_mask)
    cv2.imshow("Subtraction Result", img_bg_model)

    key = cv2.waitKey(10)
    if key == 27:
        break

错误信息 #

[ERROR:0@0.002] global /home/usr/opencv-4.x/modules/core/src/persistence.cpp (505) open Can't open file: './config/SuBSense.xml' in write mode
Failed to open ./config/SuBSense.xml

解决方案 #

• SuBSENSE算法依赖对应的XML配置文件读取运行参数，pybgs库的config文件夹中缺少该文件，需手动补充：
  1. 创建基础配置文件：直接使用以下模板生成SuBSense.xml：

     <?xml version="1.0"?>
     <opencv_storage>
     <algorithm_type>SuBSENSE</algorithm_type>
     <sampling_rate>2</sampling_rate>
     <learning_rate>0.001</learning_rate>
     <num_samples>20</num_samples>
     <lower_threshold>20</lower_threshold>
     <upper_threshold>25</upper_threshold>
     <shadow_threshold>0.5</shadow_threshold>
     </opencv_storage>
     

  2. 放置路径选择：
     • 方法一：将文件放入程序运行目录下的config文件夹（若不存在则手动创建该文件夹）。
     • 方法二：将文件放入pybgs库自带的config文件夹，同时初始化算法时指定绝对路径，示例：

       # 替换为你的pybgs config文件夹实际路径
       config_path = "/path/to/your/pybgs/config/SuBSense.xml"
       background_sub_method = bgs.SuBSENSE(config_path)
       

  3. 确保程序对目标文件夹有读写权限，避免因权限问题导致文件无法打开。

内容的提问来源于stack exchange，提问作者Sebastien Grand