脚本之家,脚本语言编程技术及教程分享平台!
分类导航

Python|VBS|Ruby|Lua|perl|VBA|Golang|PowerShell|Erlang|autoit|Dos|bat|

服务器之家 - 脚本之家 - Python - python实现单目标、多目标、多尺度、自定义特征的KCF跟踪算法(实例代码)

python实现单目标、多目标、多尺度、自定义特征的KCF跟踪算法(实例代码)

2020-05-05 11:10奔跑的Yancy Python

这篇文章主要介绍了python实现单目标、多目标、多尺度、自定义特征的KCF跟踪算法,本文通过实例代码给大家介绍的非常详细,具有一定的参考借鉴价值,需要的朋友可以参考下

单目标跟踪

直接调用opencv中封装的tracker即可。

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Jan 5 17:50:47 2020
第四章 kcf跟踪
@author: youxinlin
"""
import cv2
from items import MessageItem
import time
import numpy as np
'''
监视者模块,负责入侵检测,目标跟踪
'''
class WatchDog(object):
 #入侵检测者模块,用于入侵检测
 def __init__(self,frame=None):
  #运动检测器构造函数
  self._background = None
  if frame is not None:
   self._background = cv2.GaussianBlur(cv2.cvtColor(frame,cv2.COLOR_BGR2GRAY),(21,21),0)
  self.es = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (10, 10))
 def isWorking(self):
  #运动检测器是否工作
  return self._background is not None
 def startWorking(self,frame):
  #运动检测器开始工作
  if frame is not None:
   self._background = cv2.GaussianBlur(cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY), (21, 21), 0)
 def stopWorking(self):
  #运动检测器结束工作
  self._background = None
 def analyze(self,frame):
  #运动检测
  if frame is None or self._background is None:
   return
  sample_frame = cv2.GaussianBlur(cv2.cvtColor(frame,cv2.COLOR_BGR2GRAY),(21,21),0)
  diff = cv2.absdiff(self._background,sample_frame)
  diff = cv2.threshold(diff, 25, 255, cv2.THRESH_BINARY)[1]
  diff = cv2.dilate(diff, self.es, iterations=2)
  image, cnts, hierarchy = cv2.findContours(diff.copy(),cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
  coordinate = []
  bigC = None
  bigMulti = 0
  for c in cnts:
   if cv2.contourArea(c) < 1500:
    continue
   (x,y,w,h) = cv2.boundingRect(c)
   if w * h > bigMulti:
    bigMulti = w * h
    bigC = ((x,y),(x+w,y+h))
  if bigC:
   cv2.rectangle(frame, bigC[0],bigC[1], (255,0,0), 2, 1)
  coordinate.append(bigC)
  message = {"coord":coordinate}
  message['msg'] = None
  return MessageItem(frame,message)
class Tracker(object):
 '''
 追踪者模块,用于追踪指定目标
 '''
 def __init__(self,tracker_type = "BOOSTING",draw_coord = True):
  '''
  初始化追踪器种类
  '''
  #获得opencv版本
  (major_ver, minor_ver, subminor_ver) = (cv2.__version__).split('.')
  self.tracker_types = ['BOOSTING', 'MIL','KCF', 'TLD', 'MEDIANFLOW', 'GOTURN']
  self.tracker_type = tracker_type
  self.isWorking = False
  self.draw_coord = draw_coord
  #构造追踪器
  if int(minor_ver) < 3:
   self.tracker = cv2.Tracker_create(tracker_type)
  else:
   if tracker_type == 'BOOSTING':
    self.tracker = cv2.TrackerBoosting_create()
   if tracker_type == 'MIL':
    self.tracker = cv2.TrackerMIL_create()
   if tracker_type == 'KCF':
    self.tracker = cv2.TrackerKCF_create()
   if tracker_type == 'TLD':
    self.tracker = cv2.TrackerTLD_create()
   if tracker_type == 'MEDIANFLOW':
    self.tracker = cv2.TrackerMedianFlow_create()
   if tracker_type == 'GOTURN':
    self.tracker = cv2.TrackerGOTURN_create()
 def initWorking(self,frame,box):
  '''
  追踪器工作初始化
  frame:初始化追踪画面
  box:追踪的区域
  '''
  if not self.tracker:
   raise Exception("追踪器未初始化")
  status = self.tracker.init(frame,box)
  if not status:
   raise Exception("追踪器工作初始化失败")
  self.coord = box
  self.isWorking = True
 def track(self,frame):
  '''
  开启追踪
  '''
  message = None
  if self.isWorking:
   status,self.coord = self.tracker.update(frame)
   if status:
    message = {"coord":[((int(self.coord[0]), int(self.coord[1])),(int(self.coord[0] + self.coord[2]), int(self.coord[1] + self.coord[3])))]}
    if self.draw_coord:
     p1 = (int(self.coord[0]), int(self.coord[1]))
     p2 = (int(self.coord[0] + self.coord[2]), int(self.coord[1] + self.coord[3]))
     cv2.rectangle(frame, p1, p2, (255,0,0), 2, 1)
     message['msg'] = "is tracking"
  return MessageItem(frame,message)
class ObjectTracker(object):
 def __init__(self,dataSet):
  self.cascade = cv2.CascadeClassifier(dataSet)
 def track(self,frame):
  gray = cv2.cvtColor(frame,cv2.COLOR_BGR2GRAY)
  faces = self.cascade.detectMultiScale(gray,1.03,5)
  for (x,y,w,h) in faces:
   cv2.rectangle(frame,(x,y),(x+w,y+h),(0,0,255),2)
  return frame
if __name__ == '__main__' :
# tracker_types = ['BOOSTING', 'MIL','KCF', 'TLD', 'MEDIANFLOW', 'GOTURN']
 tracker = Tracker(tracker_type="KCF")
# video = cv2.VideoCapture(0)
# video = cv2.VideoCapture("complex1.mov")
 video = cv2.VideoCapture(r"/Users/youxinlin/Desktop/video_data/complex1.MOV")
 ok, frame = video.read()
 bbox = cv2.selectROI(frame, False)
 tracker.initWorking(frame,bbox)
 while True:
  _,frame = video.read();
  if(_):
   item = tracker.track(frame);
   cv2.imshow("track",item.getFrame())
   k = cv2.waitKey(1) & 0xff
   if k == 27:
    break

附带items.py,放在同个文件夹下:

 

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Jan 5 17:51:04 2020
@author: youxinlin
"""
import json
from utils import IOUtil
'''
信息封装类
'''
class MessageItem(object):
 #用于封装信息的类,包含图片和其他信息
 def __init__(self,frame,message):
  self._frame = frame
  self._message = message
 def getFrame(self):
  #图片信息
  return self._frame
 def getMessage(self):
  #文字信息,json格式
  return self._message
 def getBase64Frame(self):
  #返回base64格式的图片,将BGR图像转化为RGB图像
  jepg = IOUtil.array_to_bytes(self._frame[...,::-1])
  return IOUtil.bytes_to_base64(jepg)
 def getBase64FrameByte(self):
  #返回base64格式图片的bytes
  return bytes(self.getBase64Frame())
 def getJson(self):
  #获得json数据格式
  dicdata = {"frame":self.getBase64Frame().decode(),"message":self.getMessage()}
  return json.dumps(dicdata)
 def getBinaryFrame(self):
  return IOUtil.array_to_bytes(self._frame[...,::-1])

utils.py:也放在同一个文件夹下。

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Jan 5 17:51:40 2020
@author: youxinlin
"""
import time
import numpy
import base64
import os
import logging
import sys
from PIL import Image
from io import BytesIO
#工具类
class IOUtil(object):
 #流操作工具类
 @staticmethod
 def array_to_bytes(pic,formatter="jpeg",quality=70):
  '''
  静态方法,将numpy数组转化二进制流
  :param pic: numpy数组
  :param format: 图片格式
  :param quality:压缩比,压缩比越高,产生的二进制数据越短
  :return:
  '''
  stream = BytesIO()
  picture = Image.fromarray(pic)
  picture.save(stream,format=formatter,quality=quality)
  jepg = stream.getvalue()
  stream.close()
  return jepg
 @staticmethod
 def bytes_to_base64(byte):
  '''
  静态方法,bytes转base64编码
  :param byte:
  :return:
  '''
  return base64.b64encode(byte)
 @staticmethod
 def transport_rgb(frame):
  '''
  将bgr图像转化为rgb图像,或者将rgb图像转化为bgr图像
  '''
  return frame[...,::-1]
 @staticmethod
 def byte_to_package(bytes,cmd,var=1):
  '''
  将每一帧的图片流的二进制数据进行分包
  :param byte: 二进制文件
  :param cmd:命令
  :return:
  '''
  head = [ver,len(byte),cmd]
  headPack = struct.pack("!3I", *head)
  senddata = headPack+byte
  return senddata
 @staticmethod
 def mkdir(filePath):
  '''
  创建文件夹
  '''
  if not os.path.exists(filePath):
   os.mkdir(filePath)
 @staticmethod
 def countCenter(box):
  '''
  计算一个矩形的中心
  '''
  return (int(abs(box[0][0] - box[1][0])*0.5) + box[0][0],int(abs(box[0][1] - box[1][1])*0.5) +box[0][1])
 @staticmethod
 def countBox(center):
  '''
  根据两个点计算出,x,y,c,r
  '''
  return (center[0][0],center[0][1],center[1][0]-center[0][0],center[1][1]-center[0][1])
 @staticmethod
 def getImageFileName():
  return time.strftime("%Y_%m_%d_%H_%M_%S", time.localtime())+'.png'

多目标跟踪:

和单目标差不多,改用MultiTracker_create()

?
1
2
3
4
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Jan 5 18:02:33 2020

目标跟踪

 

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
@author: youxinlin
"""import numpy as np
import cv2
import sys
'''
if len(sys.argv) != 2:
 print('Input video name is missing')
 exit()
'''
print('Select multiple tracking targets')
cv2.namedWindow("tracking")
camera = cv2.VideoCapture(r"/Users/youxinlin/Desktop/video_data/complex6.MOV")
#camera = cv2.VideoCapture(0)
tracker = cv2.MultiTracker_create() #多目标跟踪
a= cv2.Tracker_c
init_once = False
ok, image=camera.read()
if not ok:
 print('Failed to read video')
 exit()
bbox1 = cv2.selectROI('tracking', image)
bbox2 = cv2.selectROI('tracking', image)
bbox3 = cv2.selectROI('tracking', image)
while camera.isOpened():
 ok, image=camera.read()
 if not ok:
  print ('no image to read')
  break
 if not init_once:
  ok = tracker.add(cv2.TrackerKCF_create(),image,bbox1)
  ok = tracker.add(cv2.TrackerKCF_create( ),image, bbox2)
  ok = tracker.add(cv2.TrackerKCF_create(),image, bbox3)
  init_once = True
 ok, boxes = tracker.update(image)
 for newbox in boxes:
  p1 = (int(newbox[0]), int(newbox[1]))
  p2 = (int(newbox[0] + newbox[2]), int(newbox[1] + newbox[3]))
  cv2.rectangle(image, p1, p2, (0,0,255))
 cv2.imshow('tracking', image)
 k = cv2.waitKey(1)
 if k == 27 : break # esc pressed

多尺度检测的KCF、自定义所用特征的KCF

在一些场景下,不想使用默认的hog特征跟踪,或需要对比不同特征的跟踪效果,那么封装好的方法似乎不可用,需要可以自己撸一波kcf的代码,从而使用自己设定的特征。

总结

以上所述是小编给大家介绍的python实现单目标、多目标、多尺度、自定义特征的KCF跟踪算法,希望对大家有所帮助,如果大家有任何疑问请给我留言,小编会及时回复大家的。在此也非常感谢大家对服务器之家网站的支持!如果你觉得本文对你有帮助,欢迎转载,烦请注明出处,谢谢!

原文链接:https://blog.csdn.net/lyxleft/article/details/103847454

延伸 · 阅读

精彩推荐