def run():
import blobs
import aruco_beta
import numpy as np
import cv2
from cv2 import aruco
import matplotlib.pyplot as plt
from ar_markers import detect_markers
control = []
for i in range(100):
control.append(i)
frame = cv2.imread("markers_test.png")
im = cv2.imread("markers_test - blank.png")
i = 0
markers, c_markers = detect_markers(frame), detect_markers(im)
marks = []
for marker in markers:
marker.highlite_marker(frame)
if i % 2 == 0:
#print(marker.id)
marks.append(marker.id)
i += 1
missing = []
xy = []
i2 = 0
for i in control:
if i in marks:
pass
else:
missing.append(i)
for i in c_markers:
if i2 % 2 == 0:
#print(int(i.id) in missing, i.id)
if int(i.id) in missing:
xy.append(i.center)
i2 += 1
x = []
y = []
for i in xy:
#print(i[0],i[1])
x.append(i[0])
y.append(i[1])
#print(xy)
return (blobs.start(x, y, xy))
def stream(self):
print('Press "q" to quit')
counter = 0
while True:
unique_markers = {}
marker_ids = []
frame = self.getMobileFrame(self.url)
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
if (counter == 5):
print(datetime.datetime.now().time())
if (self.evaluateMarkerSequence()):
print("Major diff detected!")
print()
counter = 0
markers = detect_markers(frame)
for marker in markers:
if (
marker.id not in marker_ids
): #This removes duplicate markers. Dunno why every marker is registered twice
unique_markers[marker.id] = marker.center
#unique_markers[marker.id] =(round(marker.center[0]*cm_to_pixelsX,1),round(marker.center[1]*cm_to_pixelsY,1))
marker_ids.append(marker.id)
marker.highlite_marker(frame)
self.d_list.append(unique_markers)
frame = cv2.resize(frame, (1210, 720))
cv2.imshow('Detection Frame', frame)
counter += 1
if cv2.waitKey(1) & 0xFF == ord('q'):
self.stopper.set()
break
cv2.destroyAllWindows()
def main():
motor_key = 115
while True:
frame_captured, frame = capture.read()
frame = cv2.flip(frame, -1)
frame = cv2.resize(frame, (320, 240))
#undistort
undistorted_image = undistort(frame)
#ar_marker
markers = detect_markers(undistorted_image)
for marker in markers:
marker.highlite_marker(undistorted_image)
cv2.imshow('Frame', undistorted_image)
key = cv2.waitKey(1) & 0xFF
print(key)
if key == 27:
break
elif key == ord("\t"):
captured(undistorted_image)
elif key in MOTOR_SPEEDS:
motor_key = key
motor(motor_key)
elif key == 255:
motor(motor_key)
# When everything done, release the capture
capture.release()
cv2.destroyAllWindows()
def scan(img,mnum,ccol):
global card
global cardz
markers = detect_markers(img)
for marker in markers:
marker.highlite_marker(img)
#print(marker.id)
if marker.id>=10:
colz="black"
num=int(str(marker.id)[1])
else:
colz="white"
num=int(str(marker.id))
palo=Pal(marker.id,num,marker.center,colz)
paloz=Pal(marker.id,0,0,0)
#if not (palo in card ):
if not paloz in cardz:
card.append(palo)
cardz.append(paloz)
#cv2.imshow('AR', img)
card = sorted(card, key=lambda x:x.Cen[0], reverse=False)
lens=len(cardz)
print(lens)
if lens==4:
for i in range(len(card)):
#if not card[i].Num in mnum:
mnum.append(card[i].Num)
ccol.append(card[i].colz)
if len(mnum)==4:
break
return mnum,ccol,lens
def find_mat_marker():
capture = cv2.VideoCapture(1)
# capture = cv2.VideoCapture('markcheck4.mp4')
p1x, p2x, p3x, p4x, p1y, p2y, p3y, p4y = 0, 0, 0, 0, 0, 0, 0, 0
if capture.isOpened():
frame_captured, frame = capture.read()
else:
frame_captured = False
while frame_captured:
markers = detect_markers(frame)
for marker in markers:
# print("ID {}'s position is {}".format(marker.id, marker.center))
if marker.id == 185:
p1x, p1y = marker.center[0], marker.center[1]
print("p1(x,y) : ", p1x, p1y)
elif marker.id == 992:
p2x, p2y = marker.center[0], marker.center[1]
print("p2(x,y) : ", p2x, p2y)
elif marker.id == 2808:
p3x, p3y = marker.center[0], marker.center[1]
print("p3(x,y) : ", p3x, p3y)
elif marker.id == 3540:
p4x, p4y = marker.center[0], marker.center[1]
print("p4(x,y) : ", p4x, p4y)
cv2.imshow("Test Frame", frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
frame_captured, frame = capture.read()
capture.release()
cv2.destroyAllWindows()
return p1x, p2x, p3x, p4x, p1y, p2y, p3y, p4y
def main(q):
#for capture every second
checktimeBefore = int(time.strftime('%S'))
for frame in camera.capture_continuous(rawCapture,
format="bgr",
use_video_port=True):
# grab the raw NumPy array representing the image, then initialize the timestamp
# and occupied/unoccupied text
image = frame.array
#undistort
undistorted_image = undistort(image)
#cascade
cascade(undistorted_image)
#AR marker
markers = ar_markers.detect_markers(undistorted_image)
for marker in markers:
print('marker :', marker.id) #, marker.center)
marker.highlite_marker(undistorted_image)
# show the frame
cv2.imshow("Frame", undistorted_image)
key = cv2.waitKey(1) & 0xFF
# clear the stream in preparation for the next frame
rawCapture.truncate(0)
# 1 : capture negative images every second
switch = 0
checktime = int(time.strftime('%S'))
if checktime - checktimeBefore >= 1 and switch == 1:
captured(undistorted_image)
checktimeBefore = checktime
# Threading
evt = threading.Event()
data = undistorted_image
q.put((data, evt))
# if the `q` key was pressed, break from the loop
if key == ord("q"):
break
elif key == ord("\t"):
captured(undistorted_image)
def apply(hsr, arg):
while True:
rospy.sleep(5.)
img = hsr.subscribers['head_cam'].callback.data
_, img = cv2.threshold(img, 32, 255, cv2.THRESH_BINARY)
markers = {marker.id: marker for marker in ar_markers.detect_markers(img)}.values()
for marker in markers:
marker.draw_contour(img)
cv2.imshow('head', img)
cv2.waitKey(3)
if len(markers) == 2:
return sort_markers(sum([list(marker.contours.flat) for marker in markers], []))
else:
print('found {} markers'.format(len(markers)))
rospy.sleep(1.)
def main(argv):
# 마커 인식할 이미지 파일 이름 표시
print("the image file name is " + argv)
# 이미지 파일에서 이미지 추출
frame = cv2.imread(argv, cv2.IMREAD_UNCHANGED)
# 이미지에서 marker 위치 추출(찾기)
markers = detect_markers(frame)
# 찾은 마커 개수 표시
print("{} markers found.".format(len(markers)))
# 마커 정보 표시
for marker in markers:
# id = marker.id, 마커의 중심 위치 : marker.center 표시
print("ID {}'s position is {}".format(marker.id, marker.center))
# 해당 마커의 코너를 이쁘게(?) 표시. 사각형이니 4포인트(x,y)
print("contours are :")
for pos in marker.contours:
print("\t {}".format(pos))
def main(argv):
cam_id = 0 #default 0
fps_view = False
try:
opts, args = getopt.getopt(argv, "hc:f", ["camera_id="])
except getopt.GetoptError:
print('default setting : cam id = 0, fps indication = disabled')
for opt, arg in opts:
if opt == '-h':
print('test.py -c < camera_id >f')
sys.exit()
elif opt in ("-c", "--camera"):
cam_id = int(arg)
elif opt in ("-f", "--fps"):
fps_view = True
print('Press "q" to quit')
capture = cv2.VideoCapture(cam_id)
if capture.isOpened(): #try to get the first frame
frame_captured, frame = capture.read()
else:
print('Failed to Open Camer %d' % cam_id)
frame_captured = False
while frame_captured:
frame_captured, frame = capture.read()
markers = detect_markers(frame)
for marker in markers:
marker.highlite_marker(frame)
print('Marker ID:', marker.id)
if fps_view:
putFps(frame)
cv2.imshow('Test Frame', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
#When everything done, release the capture
capture.release()
cv2.destroyAllWindows()
def do_POST(self):
self.send_response(200)
self.send_header('X-Server2Client', '123')
self.end_headers()
data = self.rfile.read(int(self.headers['Content-Length']))
data_numpy = np.asarray(bytearray(data), dtype="uint8")
img = cv2.imdecode(data_numpy, cv2.IMREAD_ANYCOLOR)
markers = detect_markers(img)
mark_result = ' '
if markers:
for marker in markers:
print("detected", marker.id)
marker.highlite_marker(img)
if marker.id == 114:
mark_result = 'left'
if marker.id == 922:
mark_result = 'right'
if marker.id == 2537:
mark_result = 'stop'
cas_result = cascade(img)
masked_img = select_white(img, 120)
k = path_decision(masked_img, 0.25)
y1, x1 = masked_img.shape
x1 = int(x1/2)
x2 = int(-k[2] * k[1] + x1)
y2 = y1-k[2]
# cv2.line(image,(x1,y1),(x2,y2),(0,0,255),2)
cv2.line(masked_img,(x1,y1),(x2,y2),(100),2)
with open('rawvideo.jpg', 'wb') as File:
File.write(data)
with open('mask.jpg', 'wb') as File:
result, masked_data = cv2.imencode('.jpg', masked_img, [int(cv2.IMWRITE_JPEG_QUALITY), 90])
File.write(masked_data)
print('m : {}'.format(k[1]))
print('forward : {}'.format(k[2]))
send = {"action": k[0], "markers" : mark_result, "cascade" : cas_result}
self.wfile.write(
bytes(json.dumps(send), encoding='utf8'))
self.wfile.write(b'\n')
def camera_callback(self,data):
#print "here"
try:
cv_image = self.bridge.imgmsg_to_cv2(data, "bgr8")
except CvBridgeError as e:
print(e)
(row,cols,channels) = cv_image.shape
im = cv_image
markers = detect_markers(im)
for marker in markers:
marker.highlite_marker(im)
print ("found")
print rospy.Time.now()
print (markers)
# display the result
cv2.imshow("img", im)
cv2.waitKey(100)
def __markers_pos(self,
num_of_markers: int,
display: bool = False) -> Dict[int, Tuple[int]]:
cap = self.capture
# Check if the camera is opened and try to get the first frame
if cap.isOpened():
frame_captured, frame = cap.read()
else:
raise MarkerRecognitionFailure()
# Init a dictionary to store positions of AR markers.
# Format: {marker's ID: marker's location}
pos = dict()
print('Vision: start markers recognition.')
while frame_captured and len(
pos
) < num_of_markers: # Break when all markers are recognized.
markers = detect_markers(frame)
for marker in markers:
if marker.id not in pos:
pos[marker.id] = tuple(marker.center)
marker.highlite_marker(frame)
# Display the feed with marker highlighting
# if the display flag is set to true.
if display:
cv2.imshow('Test Frame', frame)
# Break when 'q' is pressed (& 0xFF required for 64-bit architecture)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
else:
time.sleep(0.01)
frame_captured, frame = cap.read()
if len(pos) < num_of_markers:
raise MarkerRecognitionFailure()
return pos
def stream(self):
print('Press "q" to quit')
counter = 0
while not self.stopper.is_set():
unique_markers = {}
marker_ids = []
frame = self.getMobileFrame(self.url)
if (counter == 5):
self.evaluateMarkerSequence()
counter = 0
markers = detect_markers(frame)
for marker in markers:
if (marker.id
not in marker_ids): # This removes duplicate markers.
unique_markers[marker.id] = marker.center
marker_ids.append(marker.id)
marker.highlite_marker(frame)
self.d_list.append(unique_markers)
frame = cv2.resize(frame, (1210, 720))
cv2.imshow('Detection Frame', frame)
counter += 1
if cv2.waitKey(1) & 0xFF == ord('q'):
print("Close the kinect windows to terminate the program")
cv2.destroyAllWindows()
def marker(img):
id = 0
markers = detect_markers(img)
for marker in markers:
id=marker.id
print('detected',marker.id)
marker.highlite_marker(img)
if id == 144:
key = 'a'
self.wfile.write(bytes(json.dumps(key), encoding='utf8'))
self.wfile.write(b'\n')
'''cv2.waitKey(1)'''
time.sleep(3)#회전시간
key = 'w'
self.wfile.write(bytes(json.dumps(key), encoding='utf8'))
self.wfile.write(b'\n')
'''cv2.waitKey(1)'''
time.sleep(2)#전진시간
elif id == 922:
key = 'd'
self.wfile.write(bytes(json.dumps(key), encoding='utf8'))
self.wfile.write(b'\n')
'''cv2.waitKey(1)'''
time.sleep(3)#회전시간
key = 'w'
self.wfile.write(bytes(json.dumps(key), encoding='utf8'))
self.wfile.write(b'\n')
'''cv2.waitKey(1)'''
time.sleep(2)#전진시간
elif id == 2537:
key = 's'
else:
key = 'blank'
id=0
return key
b = bottom_right[0] - bottom_left[0]
c = bottom_left[1] - top_left[1]
d = bottom_right[1] - top_right[1]
print(a, b, c, d)
x1 = coordinates[0] / a * horizontalDist
x2 = coordinates[0] / b * horizontalDist
y1 = coordinates[1] / c * verticalDist
y2 = coordinates[1] / d * verticalDist
print(x1, x2, y1, y2)
return (int(np.mean([x1, x2])), int(np.mean([y1, y2])))
test_image = cv.imread("images/Scanned_Collin.jpg")
markers = ar.detect_markers(test_image)
print(test_image.shape)
print(markers)
for marker in markers:
marker.highlite_marker(test_image)
print()
test_point = (300, 400)
cv.circle(test_image, test_point, 10, (255, 0, 0))
point = getPaperCoordinates(test_point, markers)
print(point)
cv.circle(test_image, point, 10, (0, 255, 0))
cv.imshow('test_image', test_image)
cv.waitKey(0)
def __init__(self):
# initiliaze
rospy.init_node('GoForward', anonymous=False)
# tell user how to stop TurtleBot
rospy.loginfo("To stop TurtleBot q")
# What function to call when you ctrl + c
rospy.on_shutdown(self.shutdown)
# Create a publisher which can "talk" to TurtleBot and tell it to move
# Tip: You may need to change cmd_vel_mux/input/navi to /cmd_vel if you're not using TurtleBot2
self.cmd_vel = rospy.Publisher('cmd_vel_mux/input/navi',
Twist,
queue_size=10)
#TurtleBot will stop if we don't keep telling it to move. How often should we tell it to move? 10 HZ
r = rospy.Rate(10)
# Twist is a datatype for velocity
move_cmd = Twist()
# let's go forward at 0.2 m/s
move_cmd.linear.x = 0.3
# let's turn at 0 radians/s
#move_cmd.angular.z = 0
#let's turn at 45 deg/s
turn_cmd = Twist()
turn_cmd.linear.x = 0
turn_cmd.angular.z = radians(90)
#45 deg/s in radians/s
#let's turn at -45 deg/s
turn_cmd2 = Twist()
turn_cmd2.linear.x = 0
turn_cmd2.angular.z = radians(-90)
#45 deg/s in radians/s
#Capture the workstation video input
capture = cv2.VideoCapture(0)
if capture.isOpened(): # try to get the first frame
frame_captured, frame = capture.read()
else:
frame_captured = False
global i
i = 0
while frame_captured:
print("---------------------")
markers = detect_markers(frame)
if detect_markers(frame): # try to get the first frame
#print("detected")
for marker in markers:
print(
"111111111111111111111111111111111111111111111111111111111111111111111111111111111111"
)
marker.highlite_marker(frame)
print(marker_id)
if (marker_id == 1803):
if (i == 1):
print("break out")
continue
print(
"***************************up right**********************************"
)
i = 1
for x in range(0, 50):
self.cmd_vel.publish(move_cmd)
print("moving forward")
# wait for 0.1 seconds (10 HZ) and publish again
r.sleep()
for x in range(0, 10):
self.cmd_vel.publish(turn_cmd2)
print("ROUND MOVEEE")
r.sleep()
for x in range(0, 60):
self.cmd_vel.publish(move_cmd)
print("moving forward")
# wait for 0.1 seconds (10 HZ) and publish again
r.sleep()
elif (marker_id == 123):
print(
"***************************round and up***************************************"
)
if (i == 2):
print("break out")
continue
i = 2
for x in range(0, 25):
self.cmd_vel.publish(turn_cmd)
print("ROUND MOVEEE")
r.sleep()
for x in range(0, 80):
self.cmd_vel.publish(move_cmd)
print("moving forward")
# wait for 0.1 seconds (10 HZ) and publish again
r.sleep()
cv2.imshow('Test Frame', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
frame_captured, frame = capture.read()
# When everything done, release the capture
capture.release()
cv2.destroyAllWindows()
self.shutdown()
def main():
# construct the argument parse
parser = argparse.ArgumentParser(
description='Script to run MobileNet-SSD object detection network ')
parser.add_argument("--video", help="path to video file. If empty, camera's stream will be used")
parser.add_argument("--prototxt", default="MobileNetSSD_deploy.prototxt",
help='Path to text network file: '
'MobileNetSSD_deploy.prototxt for Caffe model or '
)
parser.add_argument("--weights", default="MobileNetSSD_deploy.caffemodel",
help='Path to weights: '
'MobileNetSSD_deploy.caffemodel for Caffe model or '
)
parser.add_argument("--thr", default=0.2, type=float, help="confidence threshold to filter out weak detections")
args = parser.parse_args()
# Labels of Network.
classNames = { 0: 'background',
1: 'aeroplane', 2: 'bicycle', 3: 'bird', 4: 'boat',
5: 'bottle', 6: 'bus', 7: 'car', 8: 'cat', 9: 'chair',
10: 'cow', 11: 'diningtable', 12: 'dog', 13: 'horse',
14: 'motorbike', 15: 'person', 16: 'pottedplant',
17: 'sheep', 18: 'sofa', 19: 'train', 20: 'tvmonitor' }
# Open video file or capture device.
if args.video:
cap = cv2.VideoCapture(args.video)
else:
cap = cv2.VideoCapture(0)
#Load the Caffe model
net = cv2.dnn.readNetFromCaffe(args.prototxt, args.weights)
while True:
#switch init
switch = 0
# Capture frame-by-frame
ret, frame = cap.read()
size = (120,90)
frame = cv2.flip(frame,-1)
frame_resized = cv2.resize(frame,size) # resize frame for prediction
frame = undistort(frame)
# MobileNet requires fixed dimensions for input image(s)
# so we have to ensure that it is resized to 300x300 pixels.
# set a scale factor to image because network the objects has differents size.
# We perform a mean subtraction (127.5, 127.5, 127.5) to normalize the input;
# after executing this command our "blob" now has the shape:
# (1, 3, 300, 300)
blob = cv2.dnn.blobFromImage(frame_resized, 0.007843, size, (127.5, 127.5, 127.5), False)
#Set to network the input blob
net.setInput(blob)
#Prediction of network
detections = net.forward()
#Size of frame resize (300x300)
cols = frame_resized.shape[1]
rows = frame_resized.shape[0]
#For get the class and location of object detected,
# There is a fix index for class, location and confidence
# value in @detections array .
for i in range(detections.shape[2]):
confidence = detections[0, 0, i, 2] #Confidence of prediction
if confidence > args.thr: # Filter prediction
class_id = int(detections[0, 0, i, 1]) # Class label
# Object location
xLeftBottom = int(detections[0, 0, i, 3] * cols)
yLeftBottom = int(detections[0, 0, i, 4] * rows)
xRightTop = int(detections[0, 0, i, 5] * cols)
yRightTop = int(detections[0, 0, i, 6] * rows)
# Factor for scale to original size of frame
heightFactor = frame.shape[0]/size[0]
widthFactor = frame.shape[1]/size[1]
# Scale object detection to frame
xLeftBottom = int(widthFactor * xLeftBottom)
yLeftBottom = int(heightFactor * yLeftBottom)
xRightTop = int(widthFactor * xRightTop)
yRightTop = int(heightFactor * yRightTop)
# Draw location of object
cv2.rectangle(frame, (xLeftBottom, yLeftBottom), (xRightTop, yRightTop),
(0, 255, 0))
# Draw label and confidence of prediction in frame resized
if class_id in classNames:
label = classNames[class_id] + ": " + str(confidence)
labelSize, baseLine = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, 0.5, 1)
yLeftBottom = max(yLeftBottom, labelSize[1])
cv2.rectangle(frame, (xLeftBottom, yLeftBottom - labelSize[1]),
(xLeftBottom + labelSize[0], yLeftBottom + baseLine),
(255, 255, 255), cv2.FILLED)
cv2.putText(frame, label, (xLeftBottom, yLeftBottom),
cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 0))
print(label) #print class and confidence
switch = 1
#AR marker
try:
markers = ar_markers.detect_markers(undistorted_image)
for marker in markers:
marker.highlite_marker(undistorted_image)
if marker.id > 0:
switch = 2
except:
pass
#Threading
evt = threading.Event()
#mobilenet
if switch = 1:
qdata = label
elif switch = 2:
qdata = marker.id
frame_captured = False
# CALIBRAZIONE TRAMITE MARKERS AR
timer = 0
found = [False, False, False, False]
markers = [
HammingMarker(0),
HammingMarker(0),
HammingMarker(0),
HammingMarker(0)
]
centers = [(0, 0), (0, 0), (0, 0), (0, 0)]
while frame_captured:
frame_orig = frame.copy() # type: object
actual_markers = detect_markers(frame)
for marker in actual_markers: # type: HammingMarker
if marker.id == 1:
markers[0] = marker
found[0] = True
centers[0] = marker.center
if marker.id == 3:
markers[1] = marker
found[1] = True
centers[1] = marker.center
if marker.id == 5:
markers[2] = marker
found[2] = True
centers[2] = marker.center
if marker.id == 7:
def detect_marker(self):
markers = detect_markers(self.img_gray)
for marker in markers:
print(marker.id)
return marker.id
return 0
def do_POST(self):
print(self.headers['X-Client2Server'])
self.send_response(200)
self.send_header('X-Server2Client', '123')
self.end_headers()
data = self.rfile.read(int(self.headers['Content-Length']))
if DISPLAY:
data = np.asarray(bytearray(data), dtype="uint8")
img = cv2.imdecode(data, cv2.IMREAD_ANYCOLOR)
img= undistort(img)
id = 0
markers = detect_markers(img)
for marker in markers:
id=marker.id
print('detected',marker.id)
marker.highlite_marker(img)
if id == 144:
result = 'a'
self.wfile.write(bytes(json.dumps(result), encoding='utf8'))
self.wfile.write(b'\n')
#cv2.waitKey(1)
sleep(3)#회전시간
result = 'w'
self.wfile.write(bytes(json.dumps(result), encoding='utf8'))
self.wfile.write(b'\n')
#cv2.waitKey(1)
sleep(2)#전진시간
elif id == 922:
result = 'd'
self.wfile.write(bytes(json.dumps(result), encoding='utf8'))
self.wfile.write(b'\n')
#cv2.waitKey(1)
sleep(3)#회전시간
result = 'w'
self.wfile.write(bytes(json.dumps(result), encoding='utf8'))
self.wfile.write(b'\n')
#cv2.waitKey(1)
sleep(2)#전진시간
elif id == 2537:
result = 's'
else:
gray_image = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
cascade_obj = objs_cascade.detectMultiScale(gray_image, scaleFactor=1.02, minNeighbors=5, minSize=(16,16))
for (x_pos, y_pos, width, height) in cascade_obj:
if(width>=40):
cv2.rectangle(img, (x_pos, y_pos), (x_pos+width, y_pos+height), (255, 255, 255), 2)
cv2.putText(img, 'Stop', (x_pos, y_pos-10), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (0, 255, 0), 2)
result="s"
self.wfile.write(bytes(json.dumps(result), encoding='utf8'))
self.wfile.write(b'\n')
#cv2.waitKey(1)
sleep(6)
result ='w'
else:
white_img = select_white(img, 160)
result, a1, a2 = set_path3(white_img,0.25)
y1, x1 = img.shape
x1 = int(x1/2)
x2 = int(-a2 * a1 + x1)
y2 = y1-a2
cv2.line(img,(x1,y1),(x2,y2),(255),2)
cv2.imshow("Processed", img)
key = result
print(key)
#data = {"action": key}
print(time(), 'Sending', data)
self.wfile.write(bytes(json.dumps(key), encoding='utf8'))
self.wfile.write(b'\n')
#cv2.imshow('image', img)
cv2.waitKey(1)
else:
with open('uploaded.jpg', 'wb') as File:
File.write(data)
print('Written to file')
def stream(self):
print('Press "q" to quit')
counter = 0
d_list = []
prev_reading = {}
current_reading = {}
while True:
#unique_markers = []
unique_markers = {}
marker_ids = []
frame = self.getMobileFrame(self.url)
#frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
if (counter == 5):
full_d = {}
for d in d_list: #d_list is a list of the 5 most recent readings
for d_key in d:
if (not d_key in full_d):
full_d[d_key] = [
] #full_d is a single dict holding all info from the 5 most recent readings.
full_d[d_key].append(d[d_key])
for d_key in full_d:
c = Counter(full_d[d_key])
value = c.most_common()[0][
0] #Get the most frequent value from the 5 most recent readings
current_reading[
d_key] = value #current_reading holds the most frequent values from the 5 most recent readings.
diff_list = []
major_diff_detected = False
for diff in list(dictdiffer.diff(prev_reading,
current_reading)):
if (diff[0] == 'change'):
if (not isInBoundary(
diff[2][0], diff[2][1],
4)): #diff in 4 pixels mean actual movement
diff_list.append((diff[1], diff[2]))
major_diff_detected = True
if (diff[0] == 'add'):
print("Added " + str(diff[2]))
if (diff[0] == 'remove'):
print("Removed " + str(diff[2]))
if (major_diff_detected):
print("Change detected on: " + str(diff_list))
prev_reading = current_reading
current_reading = {}
counter = 0
d_list = []
markers = detect_markers(frame)
for marker in markers:
if (
marker.id not in marker_ids
): #This removes duplicate markers. Dunno why every marker is registered twice
#unique_markers.append(marker)
unique_markers[marker.id] = marker.center
marker_ids.append(marker.id)
marker.highlite_marker(frame)
print()
d_list.append(unique_markers)
frame = cv2.resize(frame, (1210, 720))
cv2.imshow('Detection Frame', frame)
counter += 1
if cv2.waitKey(1) & 0xFF == ord('q'):
break
cv2.destroyAllWindows()
aligned_frames = align.process(frames)
# Get aligned frames
# aligned_depth_frame is a 640x480 depth image
aligned_depth_frame = aligned_frames.get_depth_frame()
color_frame = aligned_frames.get_color_frame()
# Validate that both frames are valid
if not aligned_depth_frame or not color_frame:
continue
depth_image = np.asanyarray(aligned_depth_frame.get_data())
color_image = np.asanyarray(color_frame.get_data())
# detect AR markers
markers = detect_markers(color_image)
for marker in markers:
marker.highlite_marker(color_image)
center = marker.get_center_cood()
img = cv2.circle(color_image, center, 5, (0, 0, 255), -1)
distance = aligned_depth_frame.get_distance(
list(center)[0],
list(center)[1])
#print('Distance : {:.3f}m'.format(distance))
marker.put_distance(color_image, round(distance, 3))
cv2.imshow('test frame', color_image)
# Remove background - Set pixels further than clipping_distance to grey
# grey_color = 153
# depth_image_3d = np.dstack((depth_image,depth_image,depth_image)) #depth image is 1 channel, color is 3 channels
def measure_length(url, user):
# construct the argument parse and parse the arguments
# ap = argparse.ArgumentParser()
# ap.add_argument("-i", "--image", required=True,
# help="path to the input image")
# ap.add_argument("-w", "--width", type=float, required=True,
# help="width of the left-most object in the image (in inches)")
# args = vars(ap.parse_args())
# load the image, convert it to grayscale, and blur it slightly
flag = True
img_url = url
print("image url :", img_url)
image = cv2.imread('./' + img_url)
markers = detect_markers(image)
marker_width = 15
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
gray = cv2.GaussianBlur(gray, (7, 7), 0)
# cv2.imshow("gaussian", gray)
# ret, gray = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
# cv2.imshow("binary", gray)
# perform edge detection, then perform a dilation + erosion to
# close gaps in between object edges
edged = cv2.Canny(gray, 50, 100)
edged = cv2.dilate(edged, None, iterations=1)
edged = cv2.erode(edged, None, iterations=1)
# find contours in the edge map
cnts = cv2.findContours(edged.copy(), cv2.RETR_EXTERNAL,
cv2.CHAIN_APPROX_SIMPLE)
cnts = imutils.grab_contours(cnts)
print(len(cnts))
# sort the contours from left-to-right and initialize the
# 'pixels per metric' calibration variable
(cnts, _) = contours.sort_contours(cnts, method="top-to-bottom")
print(len(cnts))
cnts = list(cnts)
try:
cnts.insert(0, markers[0].contours)
except Exception:
print("marker detection fail")
flag = False
return flag
# sys.exit()
cnts = tuple(cnts)
print(len(cnts))
pixelsPerMetric = None
# loop over the contours individually
result_list = list()
for idx, c in enumerate(cnts):
# if the contour is not sufficiently large, ignore it
if cv2.contourArea(c) < 100:
continue
# compute the rotated bounding box of the contour
orig = image.copy()
box = cv2.minAreaRect(c)
box = cv2.cv.BoxPoints(box) if imutils.is_cv2() else cv2.boxPoints(box)
box = np.array(box, dtype="int")
# order the points in the contour such that they appear
# in top-left, top-right, bottom-right, and bottom-left
# order, then draw the outline of the rotated bounding
# box
box = perspective.order_points(box)
cv2.drawContours(orig, [box.astype("int")], -1, (0, 255, 0), 2)
# loop over the original points and draw them
for (x, y) in box:
cv2.circle(orig, (int(x), int(y)), 5, (0, 0, 255), -1)
# unpack the ordered bounding box, then compute the midpoint
# between the top-left and top-right coordinates, followed by
# the midpoint between bottom-left and bottom-right coordinates
(tl, tr, br, bl) = box
(tltrX, tltrY) = midpoint(tl, tr)
(blbrX, blbrY) = midpoint(bl, br)
# compute the midpoint between the top-left and top-right points,
# followed by the midpoint between the top-righ and bottom-right
(tlblX, tlblY) = midpoint(tl, bl)
(trbrX, trbrY) = midpoint(tr, br)
# draw the midpoints on the image
cv2.circle(orig, (int(tltrX), int(tltrY)), 5, (255, 0, 0), -1)
cv2.circle(orig, (int(blbrX), int(blbrY)), 5, (255, 0, 0), -1)
cv2.circle(orig, (int(tlblX), int(tlblY)), 5, (255, 0, 0), -1)
cv2.circle(orig, (int(trbrX), int(trbrY)), 5, (255, 0, 0), -1)
# draw lines between the midpoints
cv2.line(orig, (int(tltrX), int(tltrY)), (int(blbrX), int(blbrY)),
(255, 0, 255), 2)
cv2.line(orig, (int(tlblX), int(tlblY)), (int(trbrX), int(trbrY)),
(255, 0, 255), 2)
# compute the Euclidean distance between the midpoints
dA = dist.euclidean((tltrX, tltrY), (blbrX, blbrY))
dB = dist.euclidean((tlblX, tlblY), (trbrX, trbrY))
# if the pixels per metric has not been initialized, then
# compute it as the ratio of pixels to supplied metric
# (in this case, inches)
if pixelsPerMetric is None:
pixelsPerMetric = dB / marker_width
# compute the size of the object
dimA = dA / pixelsPerMetric
dimB = dB / pixelsPerMetric
# draw the object sizes on the image
# show the output image
if (float(format(dimA)) > 25 or float(format(dimB)) > 25):
if idx == 0:
continue
cv2.putText(orig, "width: {:.1f}cm".format(dimB), (0, 100),
cv2.FONT_HERSHEY_SIMPLEX, 1.2, (255, 0, 0), 2)
cv2.putText(orig, "height: {:.1f}cm".format(dimA), (0, 150),
cv2.FONT_HERSHEY_SIMPLEX, 1.2, (255, 0, 0), 2)
save_path = date_upload_measured()
print('save path:' + save_path)
cv2.imwrite('./media/' + save_path, orig)
img_measured = MeasureHistory.objects.create(user_idx=user,
image=save_path,
width=dimB,
height=dimA)
img_measured.save()
img_measured.msg = '성공'
img_measured.code = 100
result_list.append(img_measured)
return result_list
from __future__ import print_function
try:
import cv2
from ar_markers import detect_markers
except ImportError:
raise Exception('Error: OpenCv is not installed')
if __name__ == '__main__':
print('Press "q" to quit')
capture = cv2.VideoCapture(0)
if capture.isOpened(): # try to get the first frame
frame_captured, frame = capture.read()
else:
frame_captured = False
while frame_captured:
markers = detect_markers(frame)
for marker in markers:
marker.highlite_marker(frame)
print(marker)
cv2.imshow('Test Frame', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
frame_captured, frame = capture.read()
# When everything done, release the capture
capture.release()
cv2.destroyAllWindows()
def main(q):
#for capture every second
checktimeBefore = int(time.strftime('%S'))
for frame in camera.capture_continuous(rawCapture,
format="bgr",
use_video_port=True):
# grab the raw NumPy array representing the image, then initialize the timestamp
# and occupied/unoccupied text
image = frame.array
#undistort
undistorted_image = undistort(image)
#--------------motor control--------------
#decision (action, round(m,4), forward, left_line, right_line, center, direction)
masked_image = select_white(undistorted_image, 160)
result = set_path3(masked_image)
#line marker
line = []
#right
for j in range(result[2]):
left_coord = (+result[5] + 1 + result[3][j], 239 - j)
line.append(left_coord)
undistorted_image = cv2.line(undistorted_image, left_coord,
left_coord, (0, 255, 0), 4)
#left
for j in range(result[2]):
right_coord = (result[5] + 1 - result[4][j], 239 - j)
line.append(right_coord)
undistorted_image = cv2.line(undistorted_image, right_coord,
right_coord, (0, 255, 0), 4)
#slope
try:
undistorted_image = cv2.line(undistorted_image, result[6][0],
result[6][1], (0, 0, 255), 4)
except:
pass
#decision motor
direction = motor(result[0], result[1])
if result[0] == 'a' or result[0] == 'd':
time.sleep(0.5)
print(result[2])
#e-stop
if result[2] > 230:
motor('s', 0)
#----------------------------
#ultrasonic
ultra = ultrasonic()
if ultra > 0 and ultra < 12:
#print('stop')
direction = motor('s', 0)
print(ultra)
#cascade
# cas = len(cascade(undistorted_image))
# if cas != 0:
# direction = motor('s')
#AR marker
markers = ar_markers.detect_markers(undistorted_image)
for marker in markers:
if marker.id == 114:
direction = motor('q', result[1])
elif marker.id == 922:
direction = motor('e', result[1])
elif marker.id == 2537:
direction = motor('s', 0)
marker.highlite_marker(undistorted_image)
#----------------------------
#putText
try:
#slope
cv2.putText(undistorted_image, 'm = ' + str(result[1]), (10, 20),
cv2.FONT_HERSHEY_SIMPLEX, 0.4, (255, 255, 255), 1)
#direction
cv2.putText(undistorted_image, direction, (10, 40),
cv2.FONT_HERSHEY_SIMPLEX, 0.4, (255, 255, 255), 1)
except:
pass
#----------------------------
# show the frame
cv2.imshow("Frame", undistorted_image)
key = cv2.waitKey(1) & 0xFF
rawCapture.truncate(0)
#Threading
if switch == 1:
evt = threading.Event()
qdata = undistorted_image
q.put((qdata, evt))
# q : break, tap : capture
if key == ord("q"):
break
elif key == ord("\t"):
captured(undistorted_image)
img = cv2.remap(img,
map1,
map2,
interpolation=cv2.INTER_LINEAR,
borderMode=cv2.BORDER_CONSTANT)
return img
while True:
frame_captured, frame = capture.read()
frame = cv2.flip(frame, -1)
frame = cv2.resize(frame, (320, 240))
#undistort
undistorted_image = frame #undistort(frame)
#ar_marker
markers = detect_markers(undistorted_image)
for marker in markers:
marker.highlite_marker(undistorted_image)
cv2.imshow('Frame', undistorted_image)
key = cv2.waitKey(1) & 0xFF
if key == ord('q'):
break
elif key == ord("\t"):
captured(undistorted_image)
# When everything done, release the capture
capture.release()
cv2.destroyAllWindows()
#!/usr/bin/env python3
import cv2
from ar_markers import detect_markers
img = cv2.imread('image-0.png')
markers = detect_markers(img)
print(markers)
for m in markers:
print(">> Found:", m.id)
from __future__ import print_function
import cv2
from ar_markers import detect_markers
if __name__ == "__main__":
print('Press "q" to quit')
camera0 = cv2.VideoCapture(0)
if camera0.isOpened(): #try to get the first frame
frame_captured, frame = camera0.read()
else:
frame_captured = False
while frame_captured:
gray_scale = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
#markers = detect_markers(frame)
markers = detect_markers(gray_scale)
print("markers:\n", markers)
for marker in markers:
marker.highlite_marker(gray_scale)
cv2.imshow('Test Frame', gray_scale)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
frame_captured, frame = camera0.read()
#When everything done, release the camera0
camera0.release()
cv2.destroyAllWindows()
def marker_search(self, args):
#tracking all global variables necessary
global whiffcount
global lastmarker
global lastmidmarker
global enemyflag
global enemybase
global carryingflag
global taglockouts
global tagtimes
global walldistance
global zdir
global ser
#get the picture
img = self.vision.get_latest_valid_picture()
#shouldnt fail, but checking if it failed anyway
if (img is not None):
#initialize counters for 400 and 200 nodes
fourMarkers = 0
twoMarkers = 0
#find the markers - puts them into an array
markers = detect_markers(img)
#initialize search for finding our location based on mid markers and other markers
highestmarker = 0
highestmidmarker = 0
#convert array of markers to array of marker IDs
mids = []
for marker in markers:
mids.append(int(marker.id))
marker.highlite_marker(img)
#for every marker id...
for mid in mids:
#if we found one in the 400's that is a location marker
if mid > 404 and mid < 500:
#increment that count
fourMarkers = fourMarkers + 1
#if this is the highest # marker we've seen, then save it
if mid % 100 > highestmarker % 100:
highestmarker = mid
#if we found one in the 200's that is a location marker
elif mid > 204 and mid < 300:
#increment that count
twoMarkers = twoMarkers + 1
#if this is the highest # marker we've seen, then save it
if mid % 100 > highestmarker % 100:
highestmarker = mid
#if we found a mid-level wall marker
elif mid < 100:
#and this is the new highest mid marker, save it
if mid > highestmidmarker:
highestmidmarker = mid
#if we found another drone (WOW)
elif (mid % 100 == 3 or mid % 100 == 4) and mid < 400:
#search for the id in taglockouts
ind = -1
for i in range(len(taglockouts)):
if taglockouts[i] == mid:
ind = i
#if we found it, check the time stamp
if ind >= 0:
#if it is too soon, skip this marker
if time.time() - tagtimes[ind] < 2:
continue
#otherwise remove it and keep going
taglockouts.remove(mid)
tagtimes.remove(tagtimes[ind])
#add the score to the ledger
f = open(scoreledger, 'a+')
f.write("found enemy " + str(mid) + ": +1 points " +
str(time.time()) + "\n")
#and add the marker to the lockouts
taglockouts.append(mid)
tagtimes.append(time.time())
#if we found our teammate (EVEN MORE WOW)
elif (mid == 403):
#write that to the ledger too, i guess
f = open(scoreledger, 'a+')
f.write("tagged teammate\n")
#if we found the enemy flag (!!!!!!!)
elif mid % 100 == 2 and mid < 402:
#save location, play noise, and grab the flag
path = os.path.abspath('\FARTNoises\FART2.wav')
os.system('start %s' % path).play()
carryingflag = True
#if by some miracle this isnt the first time we grab the flag, we dont need to save the location
if enemyflag == -1 and lastmarker != 999:
enemyflag = lastmarker
#write results to file
with open('flagloc.pk', 'wb') as f:
pickle.dump(enemyflag, f)
#if we found our base
elif mid == 401:
#...and we have the flag
if carryingflag:
#SCORE! JACKPOT BABY
carryingflag = False
f = open(scoreledger, 'a+')
f.write("flag captured: +10 points " +
str(time.time()) + "\n")
path = os.path.abspath('\FARTNoises\FART3.wav')
os.system('start %s' % path).play()
#if we found the enemy base
elif mid % 100 == 1:
#search for the id in taglockouts
ind = -1
for i in range(len(taglockouts)):
if taglockouts[i] == mid:
ind = i
#if we found it, check the time stamp
if ind >= 0:
#if it is too soon, skip this marker
if time.time() - tagtimes[ind] < 30:
continue
#otherwise remove it and keep going
else:
taglockouts.remove(mid)
tagtimes.remove(tagtimes[ind])
#add the score
f = open(scoreledger, 'a+')
f.write("enemy base tagged: +2 points " +
str(time.time()) + "\n")
#and add the marker to the lockouts
taglockouts.append(mid)
tagtimes.append(time.time())
#so now that we parsed all the markers, we gotta figure out what to do with that info
#if we found way more 400 markers than 200 markers
if fourMarkers > twoMarkers + 1:
#go down some
zdir = 0
#if we found way more 200 markers than 400 markers
elif twoMarkers > fourMarkers + 1:
#go up some
zdir = 0
#otherwise, keep it level (hopefully)
else:
zdir = 0
#if we found a highestmarker, then we have our location
#this wont happen every time, especially if we find no markers (which is not unlikely)
if highestmarker != 0:
#save it
lastmarker = highestmarker
#initialize centerpoints
p1 = 0
p2 = 0
#search for the centers of this id and this id-1
for m in markers:
if int(m.id) == lastmarker:
p1 = m.center
elif int(m.id) == lastmarker - 1:
p2 = m.center
#if we found one, then calculate wall distance
if p2 != 0:
dist = math.sqrt((int(p1[0]) - int(p2[0]))**2 +
(int(p1[1]) - int(p2[1]))**2)
walldistance = wallDistance(dist)
#print location for testing purposes
print("location: " + str(lastmarker) + "midloc: " +
str(lastmidmarker) + "walldist: " + str(walldistance))
#tell other drone where we are and where enemyflag is
stng = str(lastmarker) + "," + str(enemyflag) + "\n"
ser.write(str.encode(stng))
#read in and parse info from other team
serin = ser.readline()
serial_input = serin.decode('utf-8')
coord = serial_input.split(",")
# if serial reads correctly and gets enemy flag location from other team
if (len(coord) == 2):
print("transmission: " + str(coord[0]) + "," + str(coord[1]))
#if other team is close to us move forward a bit to avoid crashing
if ((int(coord[0]) - lastmarker) < 5):
mambo.fly_direct(roll=0,
pitch=7,
yaw=0,
vertical_movement=0,
duration=.1)
#if flag is still not found set it from other team
if enemyflag == -1:
if (coord[1] != '-'):
enemyflag = int(coord[1])
#if flag is found, write to file
if enemyflag != -1:
with open('flagloc.pk', 'wb') as f:
pickle.dump(enemyflag, f)
#if we found a highestmidmarker, then we have our mid marker location
if highestmidmarker != 0:
#save it
lastmidmarker = highestmidmarker
#if we missed a lot of markers, then increment the whiff counter
if twoMarkers < 1 and fourMarkers < 1:
whiffcount = whiffcount + 1
else:
whiffcount = 0
def setAndGetPos():
#trackbar rajzolasa
cv2.namedWindow('dataSet')
cv2.createTrackbar('threshParam1', 'dataSet', 150, 255, nothing)
cv2.createTrackbar('threshParam2', 'dataSet', 250, 255, nothing)
cv2.createTrackbar('cannyParam1', 'dataSet', 150, 255, nothing)
cv2.createTrackbar('cannyParam2', 'dataSet', 200, 255, nothing)
x = 0
y = 0
w = 0
h = 0
bool=False
while True:
#trackbar adatok lekerese
threshParam1 = cv2.getTrackbarPos('threshParam1', 'dataSet')
threshParam2 = cv2.getTrackbarPos('threshParam2', 'dataSet')
cannyParam1 = cv2.getTrackbarPos('cannyParam1', 'dataSet')
cannyParam2 = cv2.getTrackbarPos('cannyParam2', 'dataSet')
cropped=""
#uj kep
ret, LiveImg = cap.read()
#elforgatas szoge
rot=0
#marker kereses
try:
markers = detect_markers(LiveImg)
except:
continue
#ha van marker es 3663 akkor
for marker in markers:
if (marker.id == 3663):
#szog es pozicio eltarolasa
rot = Angle(marker.contours[0][0], marker.contours[1][0])
if(rot>45):
rot=rot-90
x, y, w, h = cv2.boundingRect(marker.contours) # good
#kep vagasa az elonezeti kephez
cropped = LiveImg[y:y + h, x:x + w] # good
offset = 50
#pozico ellenorzes
if (marker.contours[0][0][0] - offset < marker.contours[1][0][0] and marker.contours[0][0][0] + offset >
marker.contours[1][0][0] and marker.contours[1][0][1] - offset < marker.contours[2][0][1] and
marker.contours[1][0][1] + offset > marker.contours[2][0][1]):
bool=True
#data rajzolsa, ezt majd megoldom mashogy
LiveImg = cv2.resize(LiveImg, (640, 480), interpolation=cv2.INTER_AREA)
ret, thresh1 = cv2.threshold(LiveImg, threshParam1,threshParam2, cv2.THRESH_BINARY)
edge = cv2.Canny(thresh1, cannyParam1, cannyParam2)
edge = cv2.resize(edge, (640, 480), interpolation=cv2.INTER_AREA)
edge = cv2.cvtColor(edge, cv2.COLOR_GRAY2RGB)
try:
cropped = cv2.resize(cropped, (640, 480), interpolation=cv2.INTER_AREA)
cv2.imshow("data", np.vstack((np.hstack((LiveImg,thresh1)),np.hstack((edge, cropped)))))
except :
tmp = np.zeros((height, width, 3), np.uint8)
tmp = cv2.resize(tmp, (640, 480), interpolation=cv2.INTER_AREA)
cv2.imshow("data", np.vstack((np.hstack((LiveImg, thresh1)), np.hstack((edge, tmp)))))
if (cv2.waitKey(1) & 0xFF == ord('q') or bool==True):
#eltolas
o=25
x=x+o
y=y+o
w=w-o*2
h=h-o*2
return x, y, w, h,rot,threshParam1,threshParam2,cannyParam1,cannyParam2