def gen_mask(frec, tpoly):
mask = cv.CreateMat(int(np.round(frec.height())),
int(np.round(frec.width())), cv.CV_8UC1)
cv.Zero(mask)
pts = tuple((int(np.round(x - frec.left)), int(np.round(y - frec.top)))
for (x, y) in tpoly.contour(0))
cv.FillPoly(mask, [pts], cv.Scalar(1), lineType=8, shift=0)
return np.asarray(mask)
def drawfilledpolygon():
cv.FillPoly(img,
[[(50, 50),
(100, 100)], [(120, 20),
(150, 30)], [(50, 50),
(120, 20)], [(100, 100),
(150, 30)]],
(255, 255, 255), 4, 1)
display(img, "Destination")
cv.WaitKey(0)
def test_2a_MemCreated(self):
cnt = 0
v = []
for i in range(self.repeat):
image = cv.CreateImage(self.size, cv.IPL_DEPTH_8U, 1)
cv.FillPoly(image, [[(0, 0), (0, 100), (100, 0)]], 0)
cnt += cv.CountNonZero(image)
v.append(image)
self.assertEqual(
cnt,
0,
msg="Memorized images are not black. Mean CountNonZero=%.3f" %
(1. * cnt / self.repeat))
def run(self):
# Capture first frame to get size
frame = cv.QueryFrame(self.capture)
#nframes =+ 1
frame_size = cv.GetSize(frame)
color_image = cv.CreateImage(cv.GetSize(frame), 8, 3)
grey_image = cv.CreateImage(cv.GetSize(frame), cv.IPL_DEPTH_8U, 1)
moving_average = cv.CreateImage(cv.GetSize(frame), cv.IPL_DEPTH_32F, 3)
def totuple(a):
try:
return tuple(totuple(i) for i in a)
except TypeError:
return a
first = True
while True:
closest_to_left = cv.GetSize(frame)[0]
closest_to_right = cv.GetSize(frame)[1]
color_image = cv.QueryFrame(self.capture)
# Smooth to get rid of false positives
cv.Smooth(color_image, color_image, cv.CV_GAUSSIAN, 3, 0)
if first:
difference = cv.CloneImage(color_image)
temp = cv.CloneImage(color_image)
cv.ConvertScale(color_image, moving_average, 1.0, 0.0)
first = False
else:
cv.RunningAvg(color_image, moving_average, .1, None)
cv.ShowImage("BG", moving_average)
# Convert the scale of the moving average.
cv.ConvertScale(moving_average, temp, 1, 0.0)
# Minus the current frame from the moving average.
cv.AbsDiff(color_image, temp, difference)
#cv.ShowImage("BG",difference)
# Convert the image to grayscale.
cv.CvtColor(difference, grey_image, cv.CV_RGB2GRAY)
cv.ShowImage("BG1", grey_image)
# Convert the image to black and white.
cv.Threshold(grey_image, grey_image, 40, 255, cv.CV_THRESH_BINARY)
#cv.ShowImage("BG2", grey_image)
# Dilate and erode to get people blobs
cv.Dilate(grey_image, grey_image, None, 8)
cv.Erode(grey_image, grey_image, None, 3)
cv.ShowImage("BG3", grey_image)
storage = cv.CreateMemStorage(0)
global contour
contour = cv.FindContours(grey_image, storage, cv.CV_RETR_CCOMP,
cv.CV_CHAIN_APPROX_SIMPLE)
points = []
while contour:
global bound_rect
bound_rect = cv.BoundingRect(list(contour))
polygon_points = cv.ApproxPoly(list(contour), storage,
cv.CV_POLY_APPROX_DP)
contour = contour.h_next()
global pt1, pt2
pt1 = (bound_rect[0], bound_rect[1])
pt2 = (bound_rect[0] + bound_rect[2],
bound_rect[1] + bound_rect[3])
#size control
if (bound_rect[0] - bound_rect[2] >
10) and (bound_rect[1] - bound_rect[3] > 10):
points.append(pt1)
points.append(pt2)
#points += list(polygon_points)
global box, box2, box3, box4, box5
box = cv.MinAreaRect2(polygon_points)
box2 = cv.BoxPoints(box)
box3 = np.int0(np.around(box2))
box4 = totuple(box3)
box5 = box4 + (box4[0], )
cv.FillPoly(grey_image, [
list(polygon_points),
], cv.CV_RGB(255, 255, 255), 0, 0)
cv.PolyLine(color_image, [
polygon_points,
], 0, cv.CV_RGB(255, 255, 255), 1, 0, 0)
cv.PolyLine(color_image, [list(box5)], 0, (0, 0, 255), 2)
#cv.Rectangle(color_image, pt1, pt2, cv.CV_RGB(255,0,0), 1)
if len(points):
#center_point = reduce(lambda a, b: ((a[0] + b[0]) / 2, (a[1] + b[1]) / 2), points)
center1 = (pt1[0] + pt2[0]) / 2
center2 = (pt1[1] + pt2[1]) / 2
#print center1, center2, center_point
#cv.Circle(color_image, center_point, 40, cv.CV_RGB(255, 255, 255), 1)
#cv.Circle(color_image, center_point, 30, cv.CV_RGB(255, 100, 0), 1)
#cv.Circle(color_image, center_point, 20, cv.CV_RGB(255, 255, 255), 1)
cv.Circle(color_image, (center1, center2), 5,
cv.CV_RGB(0, 0, 255), -1)
cv.ShowImage("Target", color_image)
# Listen for ESC key
c = cv.WaitKey(7) % 0x100
if c == 27:
#cv.DestroyAllWindows()
break
def run(self):
# Initialize
# log_file_name = "tracker_output.log"
# log_file = file( log_file_name, 'a' )
print "hello"
frame = cv.QueryFrame(self.capture)
frame_size = cv.GetSize(frame)
# Capture the first frame from webcam for image properties
display_image = cv.QueryFrame(self.capture)
# Greyscale image, thresholded to create the motion mask:
grey_image = cv.CreateImage(cv.GetSize(frame), cv.IPL_DEPTH_8U, 1)
# The RunningAvg() function requires a 32-bit or 64-bit image...
running_average_image = cv.CreateImage(cv.GetSize(frame),
cv.IPL_DEPTH_32F, 3)
# ...but the AbsDiff() function requires matching image depths:
running_average_in_display_color_depth = cv.CloneImage(display_image)
# RAM used by FindContours():
mem_storage = cv.CreateMemStorage(0)
# The difference between the running average and the current frame:
difference = cv.CloneImage(display_image)
target_count = 1
last_target_count = 1
last_target_change_t = 0.0
k_or_guess = 1
codebook = []
frame_count = 0
last_frame_entity_list = []
t0 = time.time()
# For toggling display:
image_list = ["camera", "difference", "threshold", "display", "faces"]
image_index = 3 # Index into image_list
# Prep for text drawing:
text_font = cv.InitFont(cv.CV_FONT_HERSHEY_COMPLEX, .5, .5, 0.0, 1,
cv.CV_AA)
text_coord = (5, 15)
text_color = cv.CV_RGB(255, 255, 255)
# Set this to the max number of targets to look for (passed to k-means):
max_targets = 5
while True:
# Capture frame from webcam
camera_image = cv.QueryFrame(self.capture)
frame_count += 1
frame_t0 = time.time()
# Create an image with interactive feedback:
display_image = cv.CloneImage(camera_image)
# Create a working "color image" to modify / blur
color_image = cv.CloneImage(display_image)
# Smooth to get rid of false positives
cv.Smooth(color_image, color_image, cv.CV_GAUSSIAN, 19, 0)
# Use the Running Average as the static background
# a = 0.020 leaves artifacts lingering way too long.
# a = 0.320 works well at 320x240, 15fps. (1/a is roughly num frames.)
cv.RunningAvg(color_image, running_average_image, 0.320, None)
# cv.ShowImage("background ", running_average_image)
# Convert the scale of the moving average.
cv.ConvertScale(running_average_image,
running_average_in_display_color_depth, 1.0, 0.0)
# Subtract the current frame from the moving average.
cv.AbsDiff(color_image, running_average_in_display_color_depth,
difference)
cv.ShowImage("difference ", difference)
# Convert the image to greyscale.
cv.CvtColor(difference, grey_image, cv.CV_RGB2GRAY)
# Threshold the image to a black and white motion mask:
cv.Threshold(grey_image, grey_image, 2, 255, cv.CV_THRESH_BINARY)
# Smooth and threshold again to eliminate "sparkles"
cv.Smooth(grey_image, grey_image, cv.CV_GAUSSIAN, 19, 0)
cv.Threshold(grey_image, grey_image, 240, 255, cv.CV_THRESH_BINARY)
grey_image_as_array = numpy.asarray(cv.GetMat(grey_image))
non_black_coords_array = numpy.where(grey_image_as_array > 3)
# Convert from numpy.where()'s two separate lists to one list of (x, y) tuples:
non_black_coords_array = zip(non_black_coords_array[1],
non_black_coords_array[0])
points = [
] # Was using this to hold either pixel coords or polygon coords.
bounding_box_list = []
# Now calculate movements using the white pixels as "motion" data
contour = cv.FindContours(grey_image, mem_storage,
cv.CV_RETR_CCOMP,
cv.CV_CHAIN_APPROX_SIMPLE)
levels = 10
while contour:
bounding_rect = cv.BoundingRect(list(contour))
point1 = (bounding_rect[0], bounding_rect[1])
point2 = (bounding_rect[0] + bounding_rect[2],
bounding_rect[1] + bounding_rect[3])
bounding_box_list.append((point1, point2))
polygon_points = cv.ApproxPoly(list(contour), mem_storage,
cv.CV_POLY_APPROX_DP)
# To track polygon points only (instead of every pixel):
# points += list(polygon_points)
# Draw the contours:
cv.DrawContours(color_image, contour, cv.CV_RGB(255, 0, 0),
cv.CV_RGB(0, 255, 0), levels, 3, 0, (0, 0))
cv.FillPoly(grey_image, [
list(polygon_points),
], cv.CV_RGB(255, 255, 255), 0, 0)
cv.PolyLine(display_image, [
polygon_points,
], 0, cv.CV_RGB(255, 255, 255), 1, 0, 0)
# cv.Rectangle( display_image, point1, point2, cv.CV_RGB(120,120,120), 1)
contour = contour.h_next()
# Find the average size of the bbox (targets), then
# remove any tiny bboxes (which are prolly just noise).
# "Tiny" is defined as any box with 1/10th the area of the average box.
# This reduces false positives on tiny "sparkles" noise.
box_areas = []
for box in bounding_box_list:
box_width = box[right][0] - box[left][0]
box_height = box[bottom][0] - box[top][0]
box_areas.append(box_width * box_height)
# cv.Rectangle( display_image, box[0], box[1], cv.CV_RGB(255,0,0), 1)
average_box_area = 0.0
if len(box_areas):
average_box_area = float(sum(box_areas)) / len(box_areas)
trimmed_box_list = []
for box in bounding_box_list:
box_width = box[right][0] - box[left][0]
box_height = box[bottom][0] - box[top][0]
# Only keep the box if it's not a tiny noise box:
if (box_width * box_height) > average_box_area * 0.1:
trimmed_box_list.append(box)
# Draw the trimmed box list:
# for box in trimmed_box_list:
# cv.Rectangle( display_image, box[0], box[1], cv.CV_RGB(0,255,0), 2 )
bounding_box_list = merge_collided_bboxes(trimmed_box_list)
# Draw the merged box list:
for box in bounding_box_list:
cv.Rectangle(display_image, box[0], box[1],
cv.CV_RGB(0, 255, 0), 1)
# Here are our estimate points to track, based on merged & trimmed boxes:
estimated_target_count = len(bounding_box_list)
# Don't allow target "jumps" from few to many or many to few.
# Only change the number of targets up to one target per n seconds.
# This fixes the "exploding number of targets" when something stops moving
# and the motion erodes to disparate little puddles all over the place.
if frame_t0 - last_target_change_t < .350: # 1 change per 0.35 secs
estimated_target_count = last_target_count
else:
if last_target_count - estimated_target_count > 1:
estimated_target_count = last_target_count - 1
if estimated_target_count - last_target_count > 1:
estimated_target_count = last_target_count + 1
last_target_change_t = frame_t0
# Clip to the user-supplied maximum:
estimated_target_count = min(estimated_target_count, max_targets)
# The estimated_target_count at this point is the maximum number of targets
# we want to look for. If kmeans decides that one of our candidate
# bboxes is not actually a target, we remove it from the target list below.
# Using the numpy values directly (treating all pixels as points):
points = non_black_coords_array
center_points = []
if len(points):
# If we have all the "target_count" targets from last frame,
# use the previously known targets (for greater accuracy).
k_or_guess = max(estimated_target_count,
1) # Need at least one target to look for.
if len(codebook) == estimated_target_count:
k_or_guess = codebook
# points = vq.whiten(array( points )) # Don't do this! Ruins everything.
codebook, distortion = vq.kmeans(array(points), k_or_guess)
# Convert to tuples (and draw it to screen)
for center_point in codebook:
center_point = (int(center_point[0]), int(center_point[1]))
center_points.append(center_point)
# cv.Circle(display_image, center_point, 10, cv.CV_RGB(255, 0, 0), 2)
# cv.Circle(display_image, center_point, 5, cv.CV_RGB(255, 0, 0), 3)
# Now we have targets that are NOT computed from bboxes -- just
# movement weights (according to kmeans). If any two targets are
# within the same "bbox count", average them into a single target.
#
# (Any kmeans targets not within a bbox are also kept.)
trimmed_center_points = []
removed_center_points = []
for box in bounding_box_list:
# Find the centers within this box:
center_points_in_box = []
for center_point in center_points:
if center_point[0] < box[right][0] and center_point[0] > box[left][0] and \
center_point[1] < box[bottom][1] and center_point[1] > box[top][1] :
# This point is within the box.
center_points_in_box.append(center_point)
# Now see if there are more than one. If so, merge them.
if len(center_points_in_box) > 1:
# Merge them:
x_list = y_list = []
for point in center_points_in_box:
x_list.append(point[0])
y_list.append(point[1])
average_x = int(float(sum(x_list)) / len(x_list))
average_y = int(float(sum(y_list)) / len(y_list))
trimmed_center_points.append((average_x, average_y))
# Record that they were removed:
removed_center_points += center_points_in_box
if len(center_points_in_box) == 1:
trimmed_center_points.append(
center_points_in_box[0]) # Just use it.
# If there are any center_points not within a bbox, just use them.
# (It's probably a cluster comprised of a bunch of small bboxes.)
for center_point in center_points:
if (not center_point in trimmed_center_points) and (
not center_point in removed_center_points):
trimmed_center_points.append(center_point)
# Draw what we found:
# for center_point in trimmed_center_points:
# center_point = ( int(center_point[0]), int(center_point[1]) )
# cv.Circle(display_image, center_point, 20, cv.CV_RGB(255, 255,255), 1)
# cv.Circle(display_image, center_point, 15, cv.CV_RGB(100, 255, 255), 1)
# cv.Circle(display_image, center_point, 10, cv.CV_RGB(255, 255, 255), 2)
# cv.Circle(display_image, center_point, 5, cv.CV_RGB(100, 255, 255), 3)
# Determine if there are any new (or lost) targets:
actual_target_count = len(trimmed_center_points)
last_target_count = actual_target_count
# Now build the list of physical entities (objects)
this_frame_entity_list = []
# An entity is list: [ name, color, last_time_seen, last_known_coords ]
for target in trimmed_center_points:
# Is this a target near a prior entity (same physical entity)?
entity_found = False
entity_distance_dict = {}
for entity in last_frame_entity_list:
entity_coords = entity[3]
delta_x = entity_coords[0] - target[0]
delta_y = entity_coords[1] - target[1]
distance = sqrt(pow(delta_x, 2) + pow(delta_y, 2))
entity_distance_dict[distance] = entity
# Did we find any non-claimed entities (nearest to furthest):
distance_list = entity_distance_dict.keys()
distance_list.sort()
for distance in distance_list:
# Yes; see if we can claim the nearest one:
nearest_possible_entity = entity_distance_dict[distance]
# Don't consider entities that are already claimed:
if nearest_possible_entity in this_frame_entity_list:
# print "Target %s: Skipping the one iwth distance: %d at %s, C:%s" % (target, distance, nearest_possible_entity[3], nearest_possible_entity[1] )
continue
# print "Target %s: USING the one iwth distance: %d at %s, C:%s" % (target, distance, nearest_possible_entity[3] , nearest_possible_entity[1])
# Found the nearest entity to claim:
entity_found = True
nearest_possible_entity[
2] = frame_t0 # Update last_time_seen
nearest_possible_entity[
3] = target # Update the new location
this_frame_entity_list.append(nearest_possible_entity)
# log_file.write( "%.3f MOVED %s %d %d\n" % ( frame_t0, nearest_possible_entity[0], nearest_possible_entity[3][0], nearest_possible_entity[3][1] ) )
break
if entity_found == False:
# It's a new entity.
color = (random.randint(0, 255), random.randint(0, 255),
random.randint(0, 255))
name = hashlib.md5(str(frame_t0) +
str(color)).hexdigest()[:6]
last_time_seen = frame_t0
new_entity = [name, color, last_time_seen, target]
this_frame_entity_list.append(new_entity)
# log_file.write( "%.3f FOUND %s %d %d\n" % ( frame_t0, new_entity[0], new_entity[3][0], new_entity[3][1] ) )
# Now "delete" any not-found entities which have expired:
entity_ttl = 1.0 # 1 sec.
for entity in last_frame_entity_list:
last_time_seen = entity[2]
if frame_t0 - last_time_seen > entity_ttl:
# It's gone.
# log_file.write( "%.3f STOPD %s %d %d\n" % ( frame_t0, entity[0], entity[3][0], entity[3][1] ) )
pass
else:
# Save it for next time... not expired yet:
this_frame_entity_list.append(entity)
# For next frame:
last_frame_entity_list = this_frame_entity_list
# Draw the found entities to screen:
for entity in this_frame_entity_list:
center_point = entity[3]
c = entity[1] # RGB color tuple
cv.Circle(display_image, center_point, 20,
cv.CV_RGB(c[0], c[1], c[2]), 1)
cv.Circle(display_image, center_point, 15,
cv.CV_RGB(c[0], c[1], c[2]), 1)
cv.Circle(display_image, center_point, 10,
cv.CV_RGB(c[0], c[1], c[2]), 2)
cv.Circle(display_image, center_point, 5,
cv.CV_RGB(c[0], c[1], c[2]), 3)
# print "min_size is: " + str(min_size)
# Listen for ESC or ENTER key
c = cv.WaitKey(7) % 0x100
if c == 27 or c == 10:
break
# Toggle which image to show
# if chr(c) == 'd':
# image_index = ( image_index + 1 ) % len( image_list )
#
# image_name = image_list[ image_index ]
#
# # Display frame to user
# if image_name == "camera":
# image = camera_image
# cv.PutText( image, "Camera (Normal)", text_coord, text_font, text_color )
# elif image_name == "difference":
# image = difference
# cv.PutText( image, "Difference Image", text_coord, text_font, text_color )
# elif image_name == "display":
# image = display_image
# cv.PutText( image, "Targets (w/AABBs and contours)", text_coord, text_font, text_color )
# elif image_name == "threshold":
# # Convert the image to color.
# cv.CvtColor( grey_image, display_image, cv.CV_GRAY2RGB )
# image = display_image # Re-use display image here
# cv.PutText( image, "Motion Mask", text_coord, text_font, text_color )
# elif image_name == "faces":
# # Do face detection
# detect_faces( camera_image, haar_cascade, mem_storage )
# image = camera_image # Re-use camera image here
# cv.PutText( image, "Face Detection", text_coord, text_font, text_color )
# cv.ShowImage( "Target", image )
image1 = display_image
cv.ShowImage("Target 1", image1)
# if self.writer:
# cv.WriteFrame( self.writer, image );
# log_file.flush()
# If only using a camera, then there is no time.sleep() needed,
# because the camera clips us to 15 fps. But if reading from a file,
# we need this to keep the time-based target clipping correct:
frame_t1 = time.time()
# If reading from a file, put in a forced delay:
if not self.writer:
delta_t = frame_t1 - frame_t0
if delta_t < (1.0 / 15.0): time.sleep((1.0 / 15.0) - delta_t)
t1 = time.time()
time_delta = t1 - t0
processed_fps = float(frame_count) / time_delta
print "Got %d frames. %.1f s. %f fps." % (frame_count, time_delta,
processed_fps)
def segmentAndMask(img0, imgDepth0, staticMap0, mask0, thresh=10):
debug = True
staticMap = np.array(staticMap0)
imgDepth = np.array(imgDepth0)
img = np.array(img0)
mask = np.array(mask0)
#imshow('indepth', imgDepth*10)
staticMap = np.int16(staticMap)
imgDepth = np.int16(imgDepth)
noDepthMask = imgDepth == 0
#imshow('no depth mask', np.uint8(noDepthMask)*255)
diff = (imgDepth - staticMap)
adiff = np.abs(diff)
tableMask = adiff < thresh
if debug:
#imshow('adiff', adiff*15)
print imgDepth.max(), staticMap.max()
#imshow('table mask', np.uint8(tableMask)*255)
# numpy setting is really slow
maskout = (tableMask) | (mask <= 0)
maskin = (tableMask <= 0) & (mask)
cvMask = util.array2cv(np.uint8(maskout) * 255)
cvImg = util.array2cv(img)
cv.Set(cvImg, (0, 0, 0), cvMask)
img = util.cv2array(cvImg)
#img[tableMask] = 0
#img[mask <= 0] = 0
cvDep = util.array2cv(imgDepth)
depthBw = imgDepth > 0
depthBw = depthBw | noDepthMask
depthBw = depthBw & maskin
depthBw = np.uint8(depthBw) * 255
cv.Set(cvDep, 0, cvMask)
imgDepth = util.cv2array(cvDep)
imgDepth = np.uint16(imgDepth)
# Find contours. Only keep the large ones to reduce noise.
param = 2
for d in range(param):
depthBw = cv2.dilate(depthBw, None)
for d in range(param):
depthBw = cv2.erode(depthBw, None)
#imshow('depthbw', depthBw)
blobs = []
blobs = traceContoursCV2(depthBw)
blobs = [b for b in blobs if len(b) > 2]
blobs = [b for b in blobs if util.blobsize(b, ignoreLessThan=150) > 150]
foregroundMask = np.zeros(imgDepth.shape, 'uint8')
mat = util.array2cv(foregroundMask)
cv.FillPoly(mat, blobs, 255)
foregroundMask = util.cv2array(mat)
#imshow('foreground', foregroundMask)
bgMask = util.array2cv(np.uint8(foregroundMask < 0) * 255)
cv.Set(cvImg, (0, 0, 0), bgMask)
cv.Set(cvDep, 0, bgMask)
img = util.cv2array(cvImg)
imgDepth = util.cv2array(cvDep)
imgDepth = np.uint16(imgDepth)
if debug:
#imshow('seg img', .5*np.float32(img0) + .5*np.float32(img0[foregroundMask]))
img1 = img0.copy()
img1[foregroundMask <= 0] = .5 * img1[foregroundMask <= 0]
#imshow('seg img', img1)
#imshow('seg dep', imgDepth*10)
#imshow('smoothed foreground mask', foregroundMask)
return img, imgDepth, foregroundMask
for b in range(polylines_size):
pt[a][b] = (random.randrange(-width, 2 * width),
random.randrange(-height, 2 * height))
cv.PolyLine(image, pt, 1, random_color(random), random.randrange(1, 9),
line_type, 0)
cv.ShowImage(window_name, image)
cv.WaitKey(delay)
# draw some filled polylines
for i in range(number):
for a in range(nb_polylines):
for b in range(polylines_size):
pt[a][b] = (random.randrange(-width, 2 * width),
random.randrange(-height, 2 * height))
cv.FillPoly(image, pt, random_color(random), line_type, 0)
cv.ShowImage(window_name, image)
cv.WaitKey(delay)
# draw some circles
for i in range(number):
pt1 = (random.randrange(-width, 2 * width),
random.randrange(-height, 2 * height))
cv.Circle(image, pt1, random.randrange(0, 300), random_color(random),
random.randrange(-1, 9), line_type, 0)
cv.ShowImage(window_name, image)
cv.WaitKey(delay)
# draw some text
def run(self):
# Initialize
log_file_name = "tracker_output.log"
log_file = open( log_file_name, 'a' )
#fps = 25
#cap = cv2.VideoCapture( '../000104-.avi'
frame = cv.QueryFrame( self.capture )
frame_size = cv.GetSize( frame )
foreground = cv.CreateImage(cv.GetSize(frame),8,1)
foremat = cv.GetMat(foreground)
Nforemat = numpy.array(foremat, dtype=numpy.float32)
gfilter=sys.argv[2]
gfilter_string=gfilter
gfilter=float(gfilter)
print "Processing Tracker with filter: " + str(gfilter)
# Capture the first frame from webcam for image properties
display_image = cv.QueryFrame( self.capture )
# Create Background Subtractor
fgbg = cv2.BackgroundSubtractorMOG()
# Greyscale image, thresholded to create the motion mask:
grey_image = cv.CreateImage( cv.GetSize(frame), cv.IPL_DEPTH_8U, 1 )
# The RunningAvg() function requires a 32-bit or 64-bit image...
running_average_image = cv.CreateImage( cv.GetSize(frame), cv.IPL_DEPTH_32F, 3 )
# ...but the AbsDiff() function requires matching image depths:
running_average_in_display_color_depth = cv.CloneImage( display_image )
# RAM used by FindContours():
mem_storage = cv.CreateMemStorage(0)
# The difference between the running average and the current frame:
difference = cv.CloneImage( display_image )
target_count = 1
last_target_count = 1
last_target_change_t = 0.0
k_or_guess = 1
codebook=[]
frame_count=0
last_frame_entity_list = []
fps = 25
t0 = 165947
# For toggling display:
image_list = [ "camera", "shadow", "white", "threshold", "display", "yellow" ]
image_index = 0 # Index into image_list
# Prep for text drawing:
text_font = cv.InitFont(cv.CV_FONT_HERSHEY_COMPLEX, .5, .5, 0.0, 1, cv.CV_AA )
text_coord = ( 5, 15 )
text_color = cv.CV_RGB(255,255,255)
text_coord2 = ( 5, 30 )
text_coord3 = ( 5, 45 )
###############################
### Face detection stuff
#haar_cascade = cv.Load( 'haarcascades/haarcascade_frontalface_default.xml' )
#haar_cascade = cv.Load( 'C:/OpenCV2.2/data/haarcascades/haarcascade_frontalface_alt.xml' )
#haar_cascade = cv.Load( 'haarcascades/haarcascade_frontalface_alt2.xml' )
#haar_cascade = cv.Load( 'haarcascades/haarcascade_mcs_mouth.xml' )
#haar_cascade = cv.Load( 'haarcascades/haarcascade_eye.xml' )
#haar_cascade = cv.Load( 'haarcascades/haarcascade_frontalface_alt_tree.xml' )
#haar_cascade = cv.Load( 'haarcascades/haarcascade_upperbody.xml' )
#haar_cascade = cv.Load( 'haarcascades/haarcascade_profileface.xml' )
# Set this to the max number of targets to look for (passed to k-means):
max_targets = 20
while True:
# Capture frame from webcam
camera_image = cv.QueryFrame( self.capture )
#ret, frame = cap.read()
frame_count += 1
frame_t0 = time.time()
mat = cv.GetMat(camera_image)
Nmat = numpy.array(mat, dtype=numpy.uint8)
# Create an image with interactive feedback:
display_image = cv.CloneImage( camera_image )
# NEW INSERT - FGMASK
fgmask = fgbg.apply(Nmat,Nforemat,-1)
fgmask = cv.fromarray(fgmask)
# Create a working "color image" to modify / blur
color_image = cv.CloneImage( display_image )
# Smooth to get rid of false positives
cv.Smooth( color_image, color_image, cv.CV_GAUSSIAN, 19, 0 ) #Changed from 19 AND MADE MEDIAN FILTER
# Smooth to get rid of false positives
# color_image = numpy.asarray( cv.GetMat( color_image ) )
# (mu, sigma) = cv2.meanStdDev(color_image)
# edges = cv2.Canny(color_image, mu - sigma, mu + sigma)
# lines = cv2.HoughLines(edges, 1, pi / 180, 70)
# Use the Running Average as the static background
# a = 0.020 leaves artifacts lingering way too long.
# a = 0.320 works well at 320x240, 15fps. (1/a is roughly num frames.)
cv.RunningAvg( color_image, running_average_image, gfilter, None )
# Convert the scale of the moving average.
cv.ConvertScale( running_average_image, running_average_in_display_color_depth, 1.0, 0.0 )
# Subtract the current frame from the moving average.
cv.AbsDiff( color_image, running_average_in_display_color_depth, difference )
# Convert the image to greyscale.
cv.CvtColor( difference, grey_image, cv.CV_RGB2GRAY )
# Smooth Before thresholding
cv.Smooth( grey_image, grey_image, cv.CV_GAUSSIAN, 19, 19 )
# Threshold the image to a black and white motion mask:
cv.Threshold( grey_image, grey_image, 2, 255, cv.CV_THRESH_BINARY )
# Smooth and threshold again to eliminate "sparkles"
#cv.Smooth( grey_image, grey_image, cv.CV_GAUSSIAN, 19, 0 ) #changed from 19 - AND put smooth before threshold
cv.Threshold( grey_image, grey_image, 240, 255, cv.CV_THRESH_BINARY)
grey_image_as_array = numpy.asarray( cv.GetMat( grey_image ) )
non_black_coords_array = numpy.where( grey_image_as_array > 3 )
# Convert from numpy.where()'s two separate lists to one list of (x, y) tuples:
non_black_coords_array = zip( non_black_coords_array[1], non_black_coords_array[0] )
frame_hsv = cv.CreateImage(cv.GetSize(color_image),8,3)
cv.CvtColor(color_image,frame_hsv,cv.CV_BGR2HSV)
imgthreshold_yellow=cv.CreateImage(cv.GetSize(color_image),8,1)
imgthreshold_white=cv.CreateImage(cv.GetSize(color_image),8,1)
imgthreshold_white2=cv.CreateImage(cv.GetSize(color_image),8,1)
cv.InRangeS(frame_hsv,cv.Scalar(0,0,196),cv.Scalar(255,255,255),imgthreshold_white) # changed scalar from 255,15,255 to 255,255,255
cv.InRangeS(frame_hsv,cv.Scalar(41,43,224),cv.Scalar(255,255,255),imgthreshold_white2)
cv.InRangeS(frame_hsv,cv.Scalar(20,100,100),cv.Scalar(30,255,255),imgthreshold_yellow)
#cvCvtColor(color_image, yellowHSV, CV_BGR2HSV)
#lower_yellow = np.array([10, 100, 100], dtype=np.uint8)
#upper_yellow = np.array([30, 255, 255], dtype=np.uint8)
#mask_yellow = cv2.inRange(yellowHSV, lower_yellow, upper_yellow)
#res_yellow = cv2.bitwise_and(color_image, color_image, mask_yellow = mask_yellow)
points = [] # Was using this to hold either pixel coords or polygon coords.
bounding_box_list = []
# Now calculate movements using the white pixels as "motion" data
contour = cv.FindContours( grey_image, mem_storage, cv.CV_RETR_CCOMP, cv.CV_CHAIN_APPROX_SIMPLE )
i=0
while contour:
# c = contour[i]
# m = cv2.moments(c)
# Area = m['m00']
# print( Area )
bounding_rect = cv.BoundingRect( list(contour) )
point1 = ( bounding_rect[0], bounding_rect[1] )
point2 = ( bounding_rect[0] + bounding_rect[2], bounding_rect[1] + bounding_rect[3] )
bounding_box_list.append( ( point1, point2 ) )
polygon_points = cv.ApproxPoly( list(contour), mem_storage, cv.CV_POLY_APPROX_DP )
# To track polygon points only (instead of every pixel):
#points += list(polygon_points)
# Draw the contours:
###cv.DrawContours(color_image, contour, cv.CV_RGB(255,0,0), cv.CV_RGB(0,255,0), levels, 3, 0, (0,0) )
cv.FillPoly( grey_image, [ list(polygon_points), ], cv.CV_RGB(255,255,255), 0, 0 )
cv.PolyLine( display_image, [ polygon_points, ], 0, cv.CV_RGB(255,255,255), 1, 0, 0 )
#cv.Rectangle( display_image, point1, point2, cv.CV_RGB(120,120,120), 1)
# if Area > 3000:
# cv2.drawContours(imgrgb,[cnt],0,(255,255,255),2)
# print(Area)
i=i+1
contour = contour.h_next()
# Find the average size of the bbox (targets), then
# remove any tiny bboxes (which are prolly just noise).
# "Tiny" is defined as any box with 1/10th the area of the average box.
# This reduces false positives on tiny "sparkles" noise.
box_areas = []
for box in bounding_box_list:
box_width = box[right][0] - box[left][0]
box_height = box[bottom][0] - box[top][0]
box_areas.append( box_width * box_height )
#cv.Rectangle( display_image, box[0], box[1], cv.CV_RGB(255,0,0), 1)
average_box_area = 0.0
if len(box_areas): average_box_area = float( sum(box_areas) ) / len(box_areas)
trimmed_box_list = []
for box in bounding_box_list:
box_width = box[right][0] - box[left][0]
box_height = box[bottom][0] - box[top][0]
# Only keep the box if it's not a tiny noise box:
if (box_width * box_height) > average_box_area*0.1: trimmed_box_list.append( box )
# Draw the trimmed box list:
#for box in trimmed_box_list:
# cv.Rectangle( display_image, box[0], box[1], cv.CV_RGB(0,255,0), 2 )
bounding_box_list = merge_collided_bboxes( trimmed_box_list )
# Draw the merged box list:
for box in bounding_box_list:
cv.Rectangle( display_image, box[0], box[1], cv.CV_RGB(0,255,0), 1 )
# Here are our estimate points to track, based on merged & trimmed boxes:
estimated_target_count = len( bounding_box_list )
# Don't allow target "jumps" from few to many or many to few.
# Only change the number of targets up to one target per n seconds.
# This fixes the "exploding number of targets" when something stops moving
# and the motion erodes to disparate little puddles all over the place.
if frame_t0 - last_target_change_t < .35: # 1 change per 0.35 secs
estimated_target_count = last_target_count
else:
if last_target_count - estimated_target_count > 1: estimated_target_count = last_target_count - 1
if estimated_target_count - last_target_count > 1: estimated_target_count = last_target_count + 1
last_target_change_t = frame_t0
# Clip to the user-supplied maximum:
estimated_target_count = min( estimated_target_count, max_targets )
# The estimated_target_count at this point is the maximum number of targets
# we want to look for. If kmeans decides that one of our candidate
# bboxes is not actually a target, we remove it from the target list below.
# Using the numpy values directly (treating all pixels as points):
points = non_black_coords_array
center_points = []
if len(points):
# If we have all the "target_count" targets from last frame,
# use the previously known targets (for greater accuracy).
k_or_guess = max( estimated_target_count, 1 ) # Need at least one target to look for.
if len(codebook) == estimated_target_count:
k_or_guess = codebook
#points = vq.whiten(array( points )) # Don't do this! Ruins everything.
codebook, distortion = vq.kmeans( array( points ), k_or_guess )
# Convert to tuples (and draw it to screen)
for center_point in codebook:
center_point = ( int(center_point[0]), int(center_point[1]) )
center_points.append( center_point )
#cv.Circle(display_image, center_point, 10, cv.CV_RGB(255, 0, 0), 2)
#cv.Circle(display_image, center_point, 5, cv.CV_RGB(255, 0, 0), 3)
# Now we have targets that are NOT computed from bboxes -- just
# movement weights (according to kmeans). If any two targets are
# within the same "bbox count", average them into a single target.
#
# (Any kmeans targets not within a bbox are also kept.)
trimmed_center_points = []
removed_center_points = []
for box in bounding_box_list:
# Find the centers within this box:
center_points_in_box = []
for center_point in center_points:
if center_point[0] < box[right][0] and center_point[0] > box[left][0] and \
center_point[1] < box[bottom][1] and center_point[1] > box[top][1] :
# This point is within the box.
center_points_in_box.append( center_point )
# Now see if there are more than one. If so, merge them.
if len( center_points_in_box ) > 1:
# Merge them:
x_list = y_list = []
for point in center_points_in_box:
x_list.append(point[0])
y_list.append(point[1])
average_x = int( float(sum( x_list )) / len( x_list ) )
average_y = int( float(sum( y_list )) / len( y_list ) )
trimmed_center_points.append( (average_x, average_y) )
# Record that they were removed:
removed_center_points += center_points_in_box
if len( center_points_in_box ) == 1:
trimmed_center_points.append( center_points_in_box[0] ) # Just use it.
# If there are any center_points not within a bbox, just use them.
# (It's probably a cluster comprised of a bunch of small bboxes.)
for center_point in center_points:
if (not center_point in trimmed_center_points) and (not center_point in removed_center_points):
trimmed_center_points.append( center_point )
# Draw what we found:
#for center_point in trimmed_center_points:
# center_point = ( int(center_point[0]), int(center_point[1]) )
# cv.Circle(display_image, center_point, 20, cv.CV_RGB(255, 255,255), 1)
# cv.Circle(display_image, center_point, 15, cv.CV_RGB(100, 255, 255), 1)
# cv.Circle(display_image, center_point, 10, cv.CV_RGB(255, 255, 255), 2)
# cv.Circle(display_image, center_point, 5, cv.CV_RGB(100, 255, 255), 3)
# Determine if there are any new (or lost) targets:
actual_target_count = len( trimmed_center_points )
last_target_count = actual_target_count
# Now build the list of physical entities (objects)
this_frame_entity_list = []
# An entity is list: [ name, color, last_time_seen, last_known_coords ]
for target in trimmed_center_points:
# Is this a target near a prior entity (same physical entity)?
entity_found = False
entity_distance_dict = {}
for entity in last_frame_entity_list:
entity_coords= entity[3]
delta_x = entity_coords[0] - target[0]
delta_y = entity_coords[1] - target[1]
distance = sqrt( pow(delta_x,2) + pow( delta_y,2) )
entity_distance_dict[ distance ] = entity
# Did we find any non-claimed entities (nearest to furthest):
distance_list = entity_distance_dict.keys()
distance_list.sort()
for distance in distance_list:
# Yes; see if we can claim the nearest one:
nearest_possible_entity = entity_distance_dict[ distance ]
# Don't consider entities that are already claimed:
if nearest_possible_entity in this_frame_entity_list:
#print "Target %s: Skipping the one iwth distance: %d at %s, C:%s" % (target, distance, nearest_possible_entity[3], nearest_possible_entity[1] ) #Commented Out 3/20/2016
continue
#print "Target %s pixel(b,g,r) : USING the one iwth distance: %d at %s, C:%s" % (target, distance, nearest_possible_entity[3] , nearest_possible_entity[1]) # Commented Out 3/20/2016
# Found the nearest entity to claim:
entity_found = True
nearest_possible_entity[2] = frame_t0 # Update last_time_seen
nearest_possible_entity[3] = target # Update the new location
this_frame_entity_list.append( nearest_possible_entity )
#log_file.write( "%.3f MOVED %s %d %d\n" % ( frame_count, nearest_possible_entity[0], nearest_possible_entity[3][0], nearest_possible_entity[3][1] ) )
break
if entity_found == False:
# It's a new entity.
color = ( random.randint(0,255), random.randint(0,255), random.randint(0,255) )
name = hashlib.md5( str(frame_t0) + str(color) ).hexdigest()[:6]
last_time_seen = frame_t0
if imgthreshold_white[target[1],target[0]] == 0.0:
# It's a real detect (not a line marker)
new_entity = [ name, color, last_time_seen, target ]
this_frame_entity_list.append( new_entity )
log_file.write( "%.3f %.3f FOUND %s %d %d\n" % ( frame_count/fps, frame_count, new_entity[0], new_entity[3][0], new_entity[3][1] ) )
filedrive = "C:/Users/525494/New_folder/000216/run_096/"
filename = "img"+str(name)
#print "gfilter is: %.2f" + gfilter
cv.SaveImage("image-test%s-%3f.png"%(new_entity[0],gfilter), display_image)
elif imgthreshold_white[target[1],target[0]] == 255.0:
# It's a white line detect
new_entity = [ name, color, last_time_seen, target ]
this_frame_entity_list.append( new_entity )
log_file.write( "%.3f %.3f FOUND %s %d %d\n" % ( frame_count/fps, frame_count, new_entity[0], new_entity[3][0], new_entity[3][1] ) )
filedrive = "C:/Users/525494/New_folder/000216/run_096/"
filename = "img"+str(name)
#print "gfilter is: %.2f" + gfilter
cv.SaveImage("white-line-image-test%s-%3f.png"%(new_entity[0],gfilter), display_image)
elif imgthreshold_yellow[target[1],target[0]] == 255.0:
# It's a yellow line detect
new_entity = [ name, color, last_time_seen, target ]
this_frame_entity_list.append( new_entity )
log_file.write( "%.3f %.3f FOUND %s %d %d\n" % ( frame_count/fps, frame_count, new_entity[0], new_entity[3][0], new_entity[3][1] ) )
filedrive = "C:/Users/525494/New_folder/000216/run_096/"
filename = "img"+str(name)
cv.SaveImage("yellow-line-image-test%s.png"%(new_entity[0]), camera_image)
# Now "delete" any not-found entities which have expired:
entity_ttl = 1.0 # 1 sec.
for entity in last_frame_entity_list:
last_time_seen = entity[2]
if frame_t0 - last_time_seen > entity_ttl:
# It's gone.
#log_file.write( "%.3f STOPD %s %d %d\n" % ( frame_count, entity[0], entity[3][0], entity[3][1] ) )
pass
else:
# Save it for next time... not expired yet:
this_frame_entity_list.append( entity )
# For next frame:
last_frame_entity_list = this_frame_entity_list
# Draw the found entities to screen:
for entity in this_frame_entity_list:
center_point = entity[3]
c = entity[1] # RGB color tuple
cv.Circle(display_image, center_point, 20, cv.CV_RGB(c[0], c[1], c[2]), 1)
cv.Circle(display_image, center_point, 15, cv.CV_RGB(c[0], c[1], c[2]), 1)
cv.Circle(display_image, center_point, 10, cv.CV_RGB(c[0], c[1], c[2]), 2)
cv.Circle(display_image, center_point, 5, cv.CV_RGB(c[0], c[1], c[2]), 3)
#print "min_size is: " + str(min_size)
# Listen for ESC or ENTER key
c = cv.WaitKey(7) % 0x100
if c == 27 or c == 10:
break
# Toggle which image to show
if chr(c) == 'd':
image_index = ( image_index + 1 ) % len( image_list )
image_name = image_list[ image_index ]
# Display frame to user
if image_name == "camera":
image = camera_image
cv.PutText( image, "Camera (Normal)", text_coord, text_font, text_color )
elif image_name == "shadow":
image = fgmask
cv.PutText( image, "No Shadow", text_coord, text_font, text_color )
elif image_name == "white":
#image = difference
image = imgthreshold_white
cv.PutText( image, "White Threshold", text_coord, text_font, text_color )
elif image_name == "display":
#image = display_image
image = display_image
cv.PutText( image, "Targets (w/AABBs and contours)", text_coord, text_font, text_color )
cv.PutText( image, str(frame_t0), text_coord2, text_font, text_color )
cv.PutText( image, str(frame_count), text_coord3, text_font, text_color )
elif image_name == "threshold":
# Convert the image to color.
cv.CvtColor( grey_image, display_image, cv.CV_GRAY2RGB )
image = display_image # Re-use display image here
cv.PutText( image, "Motion Mask", text_coord, text_font, text_color )
elif image_name == "yellow":
# Do face detection
#detect_faces( camera_image, haar_cascade, mem_storage )
image = imgthreshold_yellow # Re-use camera image here
cv.PutText( image, "Yellow Threshold", text_coord, text_font, text_color )
#cv.ShowImage( "Target", image ) Commented out 3/19
# self.writer.write( image )
# out.write( image );
# cv.WriteFrame( self.writer, image );
# if self.writer:
# cv.WriteFrame( self.writer, image );
# video.write( image );
log_file.flush()
# If only using a camera, then there is no time.sleep() needed,
# because the camera clips us to 15 fps. But if reading from a file,
# we need this to keep the time-based target clipping correct:
frame_t1 = time.time()
# If reading from a file, put in a forced delay:
# if not self.writer:
# delta_t = frame_t1 - frame_t0
# if delta_t < ( 1.0 / 15.0 ): time.sleep( ( 1.0 / 15.0 ) - delta_t ):
if frame_count == 155740:
cv2.destroyWindow("Target")
# cv.ReleaseVideoWriter()
# self.writer.release()
# log_file.flush()
break
t1 = time.time()
time_delta = t1 - t0
processed_fps = float( frame_count ) / time_delta
print "Got %d frames. %.1f s. %f fps." % ( frame_count, time_delta, processed_fps )
