def demo_step_find_object(ip,
speed=20,
dis=1,
object='cup',
vAngle=30,
hAngle=90):
"""
移动，一步一步的寻找物体
Parameter
-------
*ip:string
-树莓派的
*object:string
-要寻找的物体
*vAngle:int
-垂直方向的角度
*hAngle:int
-水平方向的角度
*speed:int
-移动的速度
*dis:float
-时间间隔
"""
# 初始化
car = Car(ip)
# 相机初始化
camera = Camera()
camera.connect_server(ip)
camera.start_receive()
camera.thread_play()
ai = Ai(classes="./aiLib/coco/coco.names",
config="./aiLib/coco/yolov3.cfg",
weight="./aiLib/coco/yolov3.weights")
car.turn_servo_camera_vertical(vAngle)
car.turn_servo_camera_horizental(hAngle)
global STOP_FLAGE
while True:
if STOP_FLAGE is True:
print('over .............................................')
return
status = get_status_with_camera(car, camera, ai, object)
car.turn_servo_camera_horizental(90)
print(status)
if status == 'status_move':
car.run_forward(speed, dis)
if status == 'status_turn_right':
car.spin_right(10, 0.4)
if status == 'status_turn_left':
car.spin_left(10, 0.4)
if status == 'status_stop':
return
def demo_line_ai(ip, speed, dis):
"""
利用训练数据进行巡线
Param
-----
*ip: string
-树莓的ip
*speed: int
-运行速度
*dis: int
-检测一次走的距离
"""
#相机初始化
camera = Camera()
camera.connect_server(ip)
camera.start_receive()
# camera.thread_play()
#车初始化
car = Car(ip)
#Ai初始化，自己训练的结果
ai = Ai()
global STOP_FLAGE
while True:
pic = camera.take_picture() #cv2.imread("./aiLib/line/line.jpg")#
ret, name, box = ai.find_object(pic)
cv2.imshow('test', ret)
cv2.waitKey(1)
# continue
x = 0
if len(name) == 0: #没有检测到线
continue
for item in name:
if item == 'line':
x = box[0][0]
print(x, type(x))
if 250 < x < 380:
car.run_forward(speed, dis) #直行
elif x < 250:
print("spin right")
car.turn_left(speed * 0.6, dis) #左转
elif x > 380:
print("spin left")
car.turn_right(speed * 0.6, dis) #右转
if STOP_FLAGE:
return
def demo_line_algrithm(ip, speed, dis):
"""
利用图像传统知识进行巡线,对图像进行二值化，选取组最大的轮廓的中心点。图像的长640，宽480。
0 左转 250 直行 320 直行 380 右转 640
Parameter
--------
* ip:string
-树莓派的IP
* speed:int
-小车运行速度
* dis：float
-运行的时间，控制距离
"""
#相机的初始化
camera = Camera()
camera.connect_server(ip)
camera.start_receive()
camera.thread_play()
#算法初始化
algithm = Algrithm()
#小车初始化
car = Car(ip)
global STOP_FLAGE
while True:
pic = camera.take_picture()
pt = algithm.line(pic, 60)
# algithm.set_debug()
x = pt[0]
print('%d和320比较' % x)
if 250 < x < 380:
car.run_forward(speed, dis) #直行
elif x < 250:
print("spin left")
car.turn_left(speed * 0.6, dis) #左转
elif x > 380:
print("spin right")
car.turn_right(speed * 0.6, dis) #右转
if STOP_FLAGE:
return
def move_step_find_object1_thread(ip, camera, ai, object, vAngle, hAngle):
"""
小车移动寻找目标
Parameter
-------
*ip:string
-树莓派的地址
*camera：class
-相机控制
*ai:class
-人工智能类
*object:string
--目标名
*vAngle:int
--相机垂直角度
*hAngle:int
--相机水平角度
Return
---------
None
"""
global CRUSING_FLOG
global STOP_FLAGE
global CAMERA_FLAGE
car = Car(ip)
car.turn_servo_camera_vertical(vAngle)
car.turn_servo_camera_horizental(hAngle) #运动到相机指定方向
while not STOP_FLAGE:
Cruising(car, 5) # 以7的速度进行漫游
car.turn_servo_camera_vertical(vAngle - 2) #抬起相机查看
car.turn_servo_camera_horizental(hAngle)
ret = check_object_step(camera, car, ai, object) #旋转小车确认
if ret == 'find':
break
if ret == 'nothing': #没有找到，从新找
mainThread_ = threading.Thread(target=find_object,
args=(
camera,
ai,
object,
))
mainThread_.start()
CRUSING_FLOG = True
car.spin_left(2, 0.1) # 回转，减少误差
start_sensor_check = False
while True:
if STOP_FLAGE == True:
car.stop_all_wheels()
print(
'Cruising over .............................................')
return
pic = camera.take_picture()
y = []
x = []
x1 = 320
y1 = 0
for i in range(1): #多张图寻找
ret, names, box = ai.find_object(pic)
for item in names:
if item == object:
id = names.index(item)
print(box[id])
y.append(box[id][1])
x.append(box[id][0])
#多张图求平均
ysum = 0
for d in y:
ysum = ysum + d
if len(y) > 0:
y1 = ysum / len(y)
print(y1)
#多张图，求取x平均
xsum = 0
for d in x:
xsum = xsum + d
if len(x) > 0:
x1 = xsum / len(x)
print(x1)
if x1 > 360: #右转
print('turn right')
car.turn_right(4, 0.1)
elif x1 < 200: #左转
print('turn left')
car.turn_left(4, 0.1)
else: #直行
car.run_forward(4, 0.2)
if y1 > 300: #说明车子靠近目标，启动超声波，相机向下移动一点
start_sensor_check = True
car.turn_servo_camera_vertical(vAngle + 2)
#超声波检查
dis1 = 1000
dis2 = 1000
if start_sensor_check:
car.turn_servo_ultrasonic_angle(50) #左
time.sleep(0.2)
dis1 = car.distance_from_obstacle()
car.turn_servo_ultrasonic_angle(130) #右
time.sleep(0.2)
dis2 = car.distance_from_obstacle()
car.turn_servo_ultrasonic_angle(90) #中
time.sleep(0.2)
dis = car.distance_from_obstacle()
distance_to_obstacle = min(min(dis1, dis2), dis) #选区最小值
print("distance", distance_to_obstacle)
if (0 < distance_to_obstacle
and distance_to_obstacle < 30): #检查范围，满足条件说明，车找到杯子了
print("find cup")
return
def demo_move_step_find_object(ip,
speed=20,
dis=1,
object='cup',
vAngle=40,
hAngle=80):
"""
连续移动寻找物体
Parameter
-------
*ip:string
-树莓派的Ip
*object:string
-要寻找的物体
*vAngle:int
-垂直方向的角度
*hAngle:int
-水平方向的角度
"""
car = Car(ip)
camera = Camera()
camera.connect_server(ip)
camera.start_receive()
camera.thread_play()
ai = Ai(classes="./aiLib/coco/coco.names",
config="./aiLib/coco/yolov3.cfg",
weight="./aiLib/coco/yolov3.weights")
car.turn_servo_camera_vertical(vAngle)
car.turn_servo_camera_horizental(hAngle)
mainThread_ = threading.Thread(target=find_object,
args=(
camera,
ai,
object,
))
mainThread_.start()
Cruising(car, 4) # 以4的速度进行漫游
car.turn_servo_camera_vertical(vAngle)
car.turn_servo_camera_horizental(hAngle)
global STOP_FLAGE
while CAMERA_FLAGE:
if STOP_FLAGE == True:
car.stop_all_wheels()
print('over .............................................')
return
status = get_status_with_camera(car, camera, ai, object)
car.turn_servo_camera_horizental(90)
print(status)
if status == 'status_move':
car.run_forward(speed, dis)
if status == 'status_turn_right':
car.spin_right(10, 0.4)
if status == 'status_turn_left':
car.spin_left(10, 0.4)
if status == 'status_stop':
return