-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathgesture_server.py
365 lines (326 loc) · 15.5 KB
/
gesture_server.py
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
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
import socket
import web
import json
from threading import Thread,Lock
from sklearn.externals import joblib
from myConfig import *
from myGlobalVariables import *
from myCommonModules import *
from myFrameProcessing import frameProcess # 单帧处理
# 全局变量，用于REST返回,web.前缀是必须的，否则REST接口返回的都是初值
web.gGESTURE = 'invalid'# 手势名称
web.gConfidence = 1.00 # 置信度
# 全局变量互斥锁
lock = Lock()
# RESTful接口定义
urls = ('/gesture', 'RestAPI')
RestAPP = web.application(urls, globals())
class RestAPI:
def GET(self):
print('接收到REST接口手势识别结果请求，返回：')
data={}
lock.acquire()
data["gesname"]=web.gGESTURE
data["confidence"]=web.gConfidence
lock.release()
gestureJSON=json.dumps(data, cls = MyEncoder)
print(gestureJSON)
return gestureJSON
# frameBuffer为保存视频帧的全局变量，frameStates记录锁定状态，防读写冲突
# frameBuffer中保存2帧图像，socket收到数据后写入未被读锁定的变量中
# socketServer给未加读锁定（rLock）的frameBuffer元素加上写锁定（wLock），写入收到的图像，解除锁定（wOK）
# videoProcess给标记为wOK(写好)的frameBuffer元素上锁（rLock），读取一帧图像，然后解锁(rOK)
frameBuffer = [None,None]
frameStates = ['wLock','wLock']
# 将视频帧写入全局缓冲区，供videoProcess用
def writeFrameBuffer(frame):
global frameStates,frameBuffer
writeOK = True
while writeOK:
for i in [0,1]: # 读写都优先考虑frameBuffer[0]，冲突才考虑frameBuffer[1]
if frameStates[i] != 'rLock':
frameStates[i] = 'wLock'
frameBuffer[i] = np.copy(frame)
frameStates[i] = 'wOK'
writeOK = False
break
cv2.waitKey(1)
# 读取全局缓冲区中的图像
def readFrameBuffer():
global frameStates,frameBuffer
frame = None
for i in [0,1]: # 读写都优先考虑frameBuffer[0]，冲突才考虑frameBuffer[1]
if frameStates[i] == 'wOK':
frameStates[i] = 'rLock'
frame=np.copy(frameBuffer[i])
frameStates[i] = 'rOK'
break
return frame
# Socket服务器,Socket接收数据写入缓冲区，适用于多线程处理
# ip和port为侦听IP地址和端口号
# BUFSIZE是接收缓冲区大小
# BUFSIZE = 921600 # 640*480*3
# BUFSIZE = 6220800 # 1920*1080*3
# Height=480,Width=640 是视频分辨率
# ShowVideo决定是否开窗口显示收到的视频
def socketServer(ip='127.0.0.1', port=6666,BUFSIZE = None ,Height=480,Width=640,ShowVideo=False):
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1) # 重用IP和端口号
s.bind((ip,port))
s.listen(1)
print('Socket服务器端启动，开始侦听......')
if BUFSIZE is None:
BUFSIZE=Height * Width * 3 # Socket接收数据缓冲区大小
while True:
conn, addr = s.accept()
print('接到来自%s的连接' % addr[0])
while True:
# 响应键盘，等1ms，按Esc键退出
key = cv2.waitKey(1)
if key == 27: break
try:
frame = conn.recv(BUFSIZE)
r=1
while len(frame)< BUFSIZE and r<BUFSIZE:
frame += conn.recv(BUFSIZE-len(frame))
r +=1
if r>=BUFSIZE:# 循环了很多次都没有接收到数据
print('长时间接收不到客户端数据，断开连接')
s.listen(1)
print('Socket服务器端重新开始侦听......')
break
conn.send('OK'.encode('utf-8'))
except:
print('客户端断开连接')
s.listen(1)
print('Socket服务器端重新开始侦听......')
break
if len(frame) == 0: break
# print('收到视频流数据{}字节'.format(len(frame)))
frame=np.fromstring(frame, dtype='uint8')
if len(frame) == Height * Width * 3:
frame=np.reshape(frame,(Height,Width,3))
writeFrameBuffer(frame)# 写入全局缓冲区
if ShowVideo:
cv2.imshow('Server',frame)
conn.close()
# 响应键盘，等1ms，按Esc键退出
key = cv2.waitKey(1)
if key == 27:break
s.close()
return 0
# 带视频处理功能的Socket服务器,可做主程序从这里启动
# ip和port为侦听IP地址和端口号，BUFSIZE是接收缓冲区大小
# BUFSIZE = 921600 # 640*480*3
# BUFSIZE = 6220800 # 1920*1080*3
# Height=480,Width=640 是视频分辨率
# ShowVideo决定是否开窗口显示收到的视频
# hand_svm_model="hand_svm_model.m"手形识别SVM模型
# face_cascade_path,人脸检测模型路径
# moveSeg=True 启用运动分割
# saveRawVideo=None不保存原始视频，如需保存，传入保存文件路径'xxx.avi'
# saveVideo=None不保存处理过的视频，如需保存，传入保存文件路径'yyy.avi'
# collectHandData=None不采集手形数据，如需采集手形数据，传入手形名称。所采集数据供SVM训练用
def socketServerWithVideoProcess(ip='127.0.0.1', port=6666,BUFSIZE = None,Height=480,Width=640,showVideo=True,
hand_svm_model=r"..\model\hand_svm_model.m",
face_cascade_path=r"..\model\haarcascade_frontalface_alt2.xml",handSize=100,
useWaterShed=True,moveSeg=True,saveRawVideo=None,saveVideo=None,collectHandData=None):
global gGESTURE
global gConfidence
print('加载SVM手形识别模型...',end='')
handsvm = None
if not hand_svm_model is None:
if os.path.exists(hand_svm_model):
handsvm = joblib.load(hand_svm_model) # 加载训练好的手形识别svm模型
print('OK')
# 加载Haar人脸检测器
face_cascade=None
if not face_cascade_path is None:
print('加载人脸检测模型...',end='')
face_cascade = cv2.CascadeClassifier(face_cascade_path) # 加载级联分类器模型
face_cascade.load(face_cascade_path)
print('OK')
# 保存视频处理结果用
vidRaw_writer = None
vid_writer=None
if saveRawVideo:
print('创建原始视频存储对象...',end='')
saveRawVideoFileName=makeVideoFileName(filename=saveRawVideo)
vidRaw_writer = cv2.VideoWriter(saveRawVideoFileName,cv2.VideoWriter_fourcc('M','J','P','G'), 15, (Width,Height))
print('OK')
if saveVideo:
print('创建处理过的视频存储对象...',end='')
saveRecVideoFileName = makeVideoFileName(filename=saveVideo)
vid_writer = cv2.VideoWriter(saveRecVideoFileName,cv2.VideoWriter_fourcc('M','J','P','G'), 15, (Width*2,Height))
print('OK')
# 创建目录，采集手形数据图像用
if collectHandData:
if not os.path.exists(collectHandData+'_imgs'):
print('创建存储手形图像的文件夹'+collectHandData+'_imgs'+'...', end='')
os.makedirs(collectHandData+'_imgs')
print('OK')
print('启动Socket服务器...',end='')
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1) # 重用IP和端口号
s.bind((ip, port))
s.listen(1)
print('OK')
print('Socket服务器端已启动，开始侦听客户端连接......')
lastFrame = None
if BUFSIZE is None:
BUFSIZE=Height * Width * 3 # Socket接收数据缓冲区大小
while True:
conn, addr = s.accept()
print('接到来自%s的Socket连接' % addr[0])
while True:
# 响应键盘，等1ms，按Esc键退出
key = cv2.waitKey(1)
if key == 27:
print('正在关闭Socket连接...',end='')
conn.close()
print('OK')
print('正在关闭Socket服务器...',end='')
s.close()
print('OK')
if saveRawVideo: # 保存原始视频
print('正在保存原始视频...',end='')
vidRaw_writer.release()
print('OK')
if saveVideo: # 保存处理过的视频
print('正在保存处理过的视频...',end='')
vid_writer.release()
print('OK')
# 保存手形数据，采集训练数据时用
if collectHandData:
print('正在保存手形数据...', end='')
text_save(handdata, collectHandData + '_data.txt')
print('OK')
return
try:
frame = conn.recv(BUFSIZE)
r=1
while len(frame)< BUFSIZE and r<BUFSIZE:
frame += conn.recv(BUFSIZE-len(frame))
r +=1
if r>=BUFSIZE:# 循环了很多次都没有接收到数据
print('长时间接收不到客户端数据，断开连接')
s.listen(1)
print('Socket服务器端重新开始侦听......')
break
except Exception as e:
print('客户端断开连接')
s.listen(1)
print('Socket服务器端重新开始侦听......')
break
if len(frame) > 0:
# print('收到视频流数据{}字节'.format(len(frame)))
frame=np.fromstring(frame, dtype='uint8')
if len(frame) == Height * Width * 3:
frame = cv2.imdecode(frame, cv2.IMREAD_COLOR)
# frame=np.reshape(frame,(Height,Width,3)) # 如果不用编码解码，需打开此语句
frame = cv2.flip(frame, 1)
if saveRawVideo: # 保存原始视频
vidRaw_writer.write(frame)
if lastFrame is None:
lastFrame=np.copy(frame)
gesture, cg, lastFrame = frameProcess(frame, lastFrame, handsvm=handsvm, face_cascade=face_cascade,
useWaterShed=useWaterShed,moveSeg=moveSeg,handSize=handSize,
showVideo=showVideo,saveVideo=vid_writer,
collectHandData=collectHandData)
lock.acquire()
web.gGESTURE = gesture # 手势名称
web.gConfidence = cg # 置信度
lock.release()
try:
conn.send(gesture.encode('utf-8'))
except:
print('客户端断开连接')
s.listen(1)
print('Socket服务器端重新开始侦听......')
break
# 带Socket通信的视频处理程序，从Socket缓冲区获取视频，适用于多线程处理
# 每隔nFrames帧处理一次
# hand_svm_model="hand_svm_model.m"，手形识别SVM模型
# face_cascade_path="haarcascade_frontalface_alt2.xml",人脸检测模型路径
# 如果保存视频，用saveVideo=传入保存文件路径
def socketVideoProcces(nFrames=1, hand_svm_model=r"..\model\hand_svm_model.m",
face_cascade_path=r"..\model\haarcascade_frontalface_alt2.xml",saveVideo=None):
print('Socket视频处理现成已启动')
handsvm = None
if not hand_svm_model is None:
if os.path.exists(hand_svm_model):
handsvm = joblib.load(hand_svm_model)#加载训练好的手形识别svm模型
# 读入视频，提取帧图像
frame = readFrameBuffer()
while frame is None:
frame = readFrameBuffer()
# 响应键盘，等1ms，按Esc键退出
key = cv2.waitKey(1)
if key == 27: return
lastFrame = np.copy(frame) # 用np.copy(frame)比frame.copy()速度快
# 加载Haar人脸检测器
face_cascade = None
if not face_cascade_path is None:
face_cascade = cv2.CascadeClassifier(face_cascade_path) # 加载级联分类器模型
face_cascade.load(face_cascade_path)
# 保存视频处理结果用
vid_writer=None
if saveVideo:
vid_writer = cv2.VideoWriter(saveVideo,cv2.VideoWriter_fourcc('M','J','P','G'), 15, (frame.shape[1]*2,frame.shape[0]))
if nFrames <= 1: nFrames = 1
k = 0 # 记录处理过的帧数
while True:
k = (k+1) % 2592000 # 一天重置一次
t = time.time()
currentFrame = readFrameBuffer()
# print('获取视频', '成功' if not currentFrame is None else '失败')
if currentFrame is None:
cv2.waitKey(1)
continue
currentFrame = cv2.flip(currentFrame,1)
# 跳帧处理，每nFrames帧做一次处理
if k % nFrames ==0:
gesture,cg,lastFrame = frameProcess(currentFrame, lastFrame, handsvm,face_cascade,useWaterShed=True,moveSeg=True, showVideo=True,saveVideo=vid_writer)
# 响应键盘，等1ms，按Esc键退出
key = cv2.waitKey(1)
if key == 27:
# 保存手形数据，采集训练数据时用
# text_save(handdata,'data.txt')
break
if saveVideo:
vid_writer.release()
# 监测全局变量赋值情况，调试程序用
def varMonitor():
while 1:
print('监测手势识别结果：',gGESTURE)
if __name__ == '__main__': # 程序从这儿开始执行
print('启动REST接口服务...')
tRestServer = Thread(target=RestAPP.run,daemon=True) # daemon=True,线程会随着主线程退出
tRestServer.start()
# 启动监控全局变量线程，调试用
# tMonitor=Thread(target=varMonitor,daemon=True)
# tMonitor.start()
# Socket接收视频并处理，多线程
tSocketVideoProcces = Thread(target=socketServerWithVideoProcess(ip='0.0.0.0', port=VideoServerPort,
BUFSIZE=None, Height=FrameHeight, Width=FrameWidth,
showVideo=True,hand_svm_model=HandShapeSVMPath,
handSize=HandSize,
face_cascade_path=FaceCascadePath,
useWaterShed=True,moveSeg=True, saveRawVideo=None,
saveVideo=None,collectHandData=None))
tSocketVideoProcces.start()
# Socket接收视频并处理，单线程
# socketServerWithVideoProcess(ip=VideoServerIP, port=VideoServerPort, BUFSIZE=None, Height=FrameHeight,
# Width=FrameWidth, showVideo=True, hand_svm_model=HandShapeSVMPath,
# handSize=HandSize,
# face_cascade_path=FaceCascadePath,
# useWaterShed=True,moveSeg=True, saveRawVideo=None, saveVideo=None,
# collectHandData=None)
## Socket通信与Video处理多线程并行
# tSocketServer=Thread(target=socketServer)
# tSocketServer.start()
#
# tSocketVideoProcces = Thread(target=socketVideoProcces)
# tSocketVideoProcces.start()
# print('关闭REST接口服务...OK')
print('手势识别程序已退出')