deepsort问题解决记录贴

最新推荐文章于 2024-05-29 11:15:53 发布

致力于变成碰不到BUG的幸运儿

最新推荐文章于 2024-05-29 11:15:53 发布

阅读量3.7k

点赞数 1

CC 4.0 BY-SA版权

文章标签： pytorch 人工智能 python 目标跟踪

本文链接：https://blog.youkuaiyun.com/HOLEYU/article/details/121742018

在复现代码时出现问题如下：

问题一：QObject::moveToThread: Current thread (0x558247080a00) is not the object's thread (0x558247900e40).Cannot move to target thread (0x558247080a00)

QObject::moveToThread: Current thread (0x558247080a00) is not the object's thread (0x558247900e40).
Cannot move to target thread (0x558247080a00)

原因：opencv安装版本太高，需要降低版本，卸载你原来的版本，重新安装

pip install opencv-contrib-python==4.1.2.30 -i https://pypi.tuna.tsinghua.edu.cn/simple

问题二：TypeError: tuple indices must be integers or slices, not tuple

Processing frame 00002
Traceback (most recent call last):
  File "deep_sort_app.py", line 327, in <module>
    args.max_cosine_distance, args.nn_budget, args.display)
  File "deep_sort_app.py", line 270, in run
    visualizer.run(frame_callback)
  File "/home/milk/deep_sort/application_util/visualization.py", line 100, in run
    self.viewer.run(lambda: self._update_fun(frame_callback))
  File "/home/milk/deep_sort/application_util/image_viewer.py", line 305, in run
    self._terminate = not self._user_fun()
  File "/home/milk/deep_sort/application_util/visualization.py", line 100, in <lambda>
    self.viewer.run(lambda: self._update_fun(frame_callback))
  File "/home/milk/deep_sort/application_util/visualization.py", line 105, in _update_fun
    frame_callback(self, self.frame_idx)
  File "deep_sort_app.py", line 242, in frame_callback
    tracker.update(detections)
  File "/home/milk/deep_sort/deep_sort/tracker.py", line 69, in update
    self._match(detections)
  File "/home/milk/deep_sort/deep_sort/tracker.py", line 127, in _match
    detections, iou_track_candidates, unmatched_detections)
  File "/home/milk/deep_sort/deep_sort/linear_assignment.py", line 63, in min_cost_matching
    if col not in indices[:, 1]:
TypeError: tuple indices must be integers or slices, not tuple

这个出现原因相当奇怪，代码一模一样，就是突然出现的，网上看到的基本都是修改代码段，很明显问题应该不是这里，重新建立文件夹，运行源代码，又可以运行了，把不能运行的代码复制过来之后也可以运行，环境也一样，问题出现在哪？

怀疑：可能是回调文件出现问题，去看def frame_callback(vis, frame_idx)了，发现问题再回来继续写

debug之后发现不是代码的问题，是引入的库包不行，降低scikit-learn版本，卸载原来的版本，重新安装

pip install scikit-learn==0.19.2

很好，出现新的问题：

问题三：

Processing frame 00001
Processing frame 00002
Processing frame 00003
Processing frame 00004
Traceback (most recent call last):
  File "deep_sort_app.py", line 327, in <module>
    args.max_cosine_distance, args.nn_budget, args.display)
  File "deep_sort_app.py", line 270, in run
    visualizer.run(frame_callback)
  File "/home/milk/deep_sort/application_util/visualization.py", line 100, in run
    self.viewer.run(lambda: self._update_fun(frame_callback))
  File "/home/milk/deep_sort/application_util/image_viewer.py", line 305, in run
    self._terminate = not self._user_fun()
  File "/home/milk/deep_sort/application_util/visualization.py", line 100, in <lambda>
    self.viewer.run(lambda: self._update_fun(frame_callback))
  File "/home/milk/deep_sort/application_util/visualization.py", line 105, in _update_fun
    frame_callback(self, self.frame_idx)
  File "deep_sort_app.py", line 242, in frame_callback
    tracker.update(detections)
  File "/home/milk/deep_sort/deep_sort/tracker.py", line 81, in update
    self._match(detections)
  File "/home/milk/deep_sort/deep_sort/tracker.py", line 127, in _match
    self.tracks, detections, confirmed_tracks)
  File "/home/milk/deep_sort/deep_sort/linear_assignment.py", line 132, in matching_cascade
    k for k in track_indices
  File "/home/milk/deep_sort/deep_sort/linear_assignment.py", line 133, in <listcomp>
    if tracks[k].time_since_update == 1 + level
NameError: name 'tracks' is not defined

博主疯狂了，他决定不再改了，就这样吧