def jitter( self, i_image , i_n_examples ):
"""1) Apply various random affine transform to i_image (scale, rotation),
where i_n_examples is the number of transformations. The affine transforms happen
around the center of the original region of interest.
2) Translate roi with various values - crop the image from this region.
3) The same normalisation is then applied to all the cropped images, specified by setParams"""
#Store transforms applied in format tx ty scale angle
o_transforms = numpy.array([0., 0., 1., 0.])
if self.__roi == None:
print "No region of interest - returning input image!"
return None
#Always return the normalised verion of the input image
image = cv.Ipl2NumPy(self.normalise(i_image))
o_data = numpy.tile( None, (image.shape[0], image.shape[1], 1))
o_data[:,:,0] = image
if i_n_examples == 0:
return ( o_data, o_transforms)
#Rotation point should be around original roi center
center = cv.cvPoint2D32f( self.__roi.x + self.__roi.width/2,self.__roi.y + self.__roi.height/2 )
angles = numpy.random.uniform(-30.,30.,i_n_examples)
scales = numpy.random.uniform(0.9,1.1, i_n_examples)
tx = numpy.int32( numpy.round( numpy.random.uniform(-20,20,i_n_examples)))
ty = numpy.int32( numpy.round( numpy.random.uniform(-20,20,i_n_examples)))
x = self.__roi.x + tx
y = self.__roi.y + ty
#FIXME: Extend valid indices to outliers due to affine transform!!!
min_x = 0; min_y = 0;
max_x = i_image.width - self.__roi.width - 1
max_y = i_image.height - self.__roi.height - 1
valid_x = numpy.hstack([numpy.nonzero( x >= min_x )[0], numpy.nonzero( x < max_x)[0]])
valid_y = numpy.hstack([numpy.nonzero( y >= min_y )[0], numpy.nonzero( y < max_y)[0]])
valid_idx = numpy.unique(numpy.hstack([valid_x, valid_y]))
original_roi = cv.cvRect( self.__roi.x, self.__roi.y, self.__roi.width, self.__roi.height)
if self.__rot_mat == None:
original_rot_matrix = None
else:
original_rot_matrix = cv.cvCloneImage(self.__rot_mat)
for index in valid_idx:
params = numpy.array([tx[index], ty[index], scales[index], angles[index] ])
self.setAffineTransform( center, scales[index], angles[index])
self.__roi.x = int( x[index] )
self.__roi.y = int( y[index] )
image = cv.Ipl2NumPy(self.normalise( i_image ))
o_data = numpy.dstack( [o_data, image])
o_transforms = numpy.vstack( [o_transforms, params])
#Restore the original region of interest
self.__roi.x = original_roi.x
self.__roi.y = original_roi.y
if original_rot_matrix == None:
self.__rot_mat= None
else:
self.__rot_mat = cv.cvCloneImage(original_rot_matrix)
return (o_data, o_transforms)
def __init__(self, transFunc, seeds, deltas, names, pts, img, cam_proj_mat, cam_centers, zoom, id):
'''
transFunc => takes in param vector, returns homogeneous Xform
seeds => initial param vector (1XM array)
deltas => step sizes in param vector quantities (1XM array)
pts => 3D points to be mapped into img (3XN array)
img => an OpenCV image from the camera
'''
self.dataset_id = id
# Note that adaptors are deprecated...
imgTmp = cv.cvCloneImage(img)
imNP = cv.adaptors.Ipl2NumPy(imgTmp)
self.height = imNP.shape[0] * zoom
self.width = imNP.shape[1] * zoom
self.zoom = zoom
pyc.size(self.width + 200, self.height)
#pyc.size(self.height + 200, self.width + 200)
self.pts = pts
self.img = img
self.vals = seeds
self.selected = np.zeros( self.vals.shape[0] )
self.selected[0] = 1.0
self.dels = deltas
self.names = names
self.transFunc = transFunc
self.cam_proj_mat = cam_proj_mat
self.cam_centers = cam_centers
self.display_laser = True
def __init__(self, transFunc, seeds, deltas, names, pts, img, cam_proj_mat,
cam_centers, zoom, id):
'''
transFunc => takes in param vector, returns homogeneous Xform
seeds => initial param vector (1XM array)
deltas => step sizes in param vector quantities (1XM array)
pts => 3D points to be mapped into img (3XN array)
img => an OpenCV image from the camera
'''
self.dataset_id = id
# Note that adaptors are deprecated...
imgTmp = cv.cvCloneImage(img)
imNP = cv.adaptors.Ipl2NumPy(imgTmp)
self.height = imNP.shape[0] * zoom
self.width = imNP.shape[1] * zoom
self.zoom = zoom
pyc.size(self.width + 200, self.height)
#pyc.size(self.height + 200, self.width + 200)
self.pts = pts
self.img = img
self.vals = seeds
self.selected = np.zeros(self.vals.shape[0])
self.selected[0] = 1.0
self.dels = deltas
self.names = names
self.transFunc = transFunc
self.cam_proj_mat = cam_proj_mat
self.cam_centers = cam_centers
self.display_laser = True
def findcontours(iplimage, threshold=100):
srcimage = opencv.cvCloneImage(iplimage)
# create the storage area and bw image
grayscale = opencv.cvCreateImage(opencv.cvGetSize(srcimage), 8, 1)
opencv.cvCvtColor(srcimage, grayscale, opencv.CV_BGR2GRAY)
# threshold
opencv.cvThreshold(grayscale, grayscale, threshold, 255, opencv.CV_THRESH_BINARY)
storage = opencv.cvCreateMemStorage(0)
opencv.cvClearMemStorage(storage)
# find the contours
nb_contours, contours = opencv.cvFindContours(grayscale, storage)
# comment this out if you do not want approximation
contours = opencv.cvApproxPoly(contours, opencv.sizeof_CvContour, storage, opencv.CV_POLY_APPROX_DP, 3, 1)
# next line is for ctypes-opencv
# contours = opencv.cvApproxPoly (contours, opencv.sizeof(opencv.CvContour), storage, opencv.CV_POLY_APPROX_DP, 3, 1)
conts = []
for cont in contours.hrange():
points = []
for pt in cont:
points.append((pt.x, pt.y))
conts.append(points)
opencv.cvReleaseMemStorage(storage)
opencv.cvReleaseImage(srcimage)
opencv.cvReleaseImage(grayscale)
return (nb_contours, conts)
def findcontours(iplimage, threshold=100):
srcimage = opencv.cvCloneImage(iplimage)
# create the storage area and bw image
grayscale = opencv.cvCreateImage(opencv.cvGetSize(srcimage), 8, 1)
opencv.cvCvtColor(srcimage, grayscale, opencv.CV_BGR2GRAY)
#threshold
opencv.cvThreshold(grayscale, grayscale, threshold, 255, opencv.CV_THRESH_BINARY)
storage = opencv.cvCreateMemStorage(0)
opencv.cvClearMemStorage(storage)
# find the contours
nb_contours, contours = opencv.cvFindContours (grayscale, storage)
# comment this out if you do not want approximation
contours = opencv.cvApproxPoly (contours, opencv.sizeof_CvContour, storage, opencv.CV_POLY_APPROX_DP, 3, 1)
# next line is for ctypes-opencv
#contours = opencv.cvApproxPoly (contours, opencv.sizeof(opencv.CvContour), storage, opencv.CV_POLY_APPROX_DP, 3, 1)
conts = []
for cont in contours.hrange():
points=[]
for pt in cont:
points.append((pt.x,pt.y))
conts.append(points)
opencv.cvReleaseMemStorage(storage)
opencv.cvReleaseImage(srcimage)
opencv.cvReleaseImage(grayscale)
return (nb_contours, conts)
def ipl2cairo(iplimage):
srcimage = opencv.cvCloneImage(iplimage)
width = srcimage.width
height = srcimage.height
image = opencv.cvCreateImage(opencv.cvGetSize(srcimage),8, 4)
opencv.cvCvtColor(srcimage,image,opencv.CV_BGR2BGRA)
buffer = numpy.fromstring(image.imageData, dtype=numpy.uint32).astype(numpy.uint32)
buffer.shape = (image.width, image.height)
opencv.cvReleaseImage(srcimage)
opencv.cvReleaseImage(image)
return cairo.ImageSurface.create_for_data(buffer, cairo.FORMAT_RGB24, width, height, width*4)
def ipl2cairo(iplimage):
srcimage = opencv.cvCloneImage(iplimage)
width = srcimage.width
height = srcimage.height
image = opencv.cvCreateImage(opencv.cvGetSize(srcimage), 8, 4)
opencv.cvCvtColor(srcimage, image, opencv.CV_BGR2BGRA)
buffer = numpy.fromstring(image.imageData, dtype=numpy.uint32).astype(numpy.uint32)
buffer.shape = (image.width, image.height)
opencv.cvReleaseImage(srcimage)
opencv.cvReleaseImage(image)
return cairo.ImageSurface.create_for_data(buffer, cairo.FORMAT_RGB24, width, height, width * 4)
def __init__(self, frame):
self.blob_image = opencv.cvCloneImage(frame.iplimage)
self.blob_gray = opencv.cvCreateImage(opencv.cvGetSize(self.blob_image), 8, 1)
self.blob_mask = opencv.cvCreateImage(opencv.cvGetSize(self.blob_image), 8, 1)
opencv.cvSet(self.blob_mask, 1)
opencv.cvCvtColor(self.blob_image, self.blob_gray, opencv.CV_BGR2GRAY)
# opencv.cvEqualizeHist(self.blob_gray, self.blob_gray)
# opencv.cvThreshold(self.blob_gray, self.blob_gray, frame.thresh, 255, opencv.CV_THRESH_BINARY)
# opencv.cvThreshold(self.blob_gray, self.blob_gray, frame.thresh, 255, opencv.CV_THRESH_TOZERO)
opencv.cvThreshold(self.blob_gray, self.blob_gray, frame.bthresh, 255, frame.bthreshmode)
# opencv.cvAdaptiveThreshold(self.blob_gray, self.blob_gray, 255, opencv.CV_ADAPTIVE_THRESH_MEAN_C, opencv.CV_THRESH_BINARY_INV)
self._frame_blobs = CBlobResult(self.blob_gray, self.blob_mask, 100, True)
self._frame_blobs.filter_blobs(10, 10000)
self.count = self._frame_blobs.GetNumBlobs()
self.items = []
for i in range(self.count):
self.items.append(self._frame_blobs.GetBlob(i))
def __init__(self, frame):
self.blob_image = opencv.cvCloneImage(frame.iplimage)
self.blob_gray = opencv.cvCreateImage(opencv.cvGetSize(self.blob_image), 8, 1)
self.blob_mask = opencv.cvCreateImage(opencv.cvGetSize(self.blob_image), 8, 1)
opencv.cvSet(self.blob_mask, 1)
opencv.cvCvtColor(self.blob_image, self.blob_gray, opencv.CV_BGR2GRAY)
#opencv.cvEqualizeHist(self.blob_gray, self.blob_gray)
#opencv.cvThreshold(self.blob_gray, self.blob_gray, frame.thresh, 255, opencv.CV_THRESH_BINARY)
#opencv.cvThreshold(self.blob_gray, self.blob_gray, frame.thresh, 255, opencv.CV_THRESH_TOZERO)
opencv.cvThreshold(self.blob_gray, self.blob_gray, frame.bthresh, 255, frame.bthreshmode)
#opencv.cvAdaptiveThreshold(self.blob_gray, self.blob_gray, 255, opencv.CV_ADAPTIVE_THRESH_MEAN_C, opencv.CV_THRESH_BINARY_INV)
self._frame_blobs = CBlobResult(self.blob_gray, self.blob_mask, 100, True)
self._frame_blobs.filter_blobs(10, 10000)
self.count = self._frame_blobs.GetNumBlobs()
self.items = []
for i in range(self.count):
self.items.append(self._frame_blobs.GetBlob(i))
def detectHaar(iplimage, classifier):
srcimage = opencv.cvCloneImage(iplimage)
grayscale = opencv.cvCreateImage(opencv.cvGetSize(srcimage), 8, 1)
opencv.cvCvtColor(srcimage, grayscale, opencv.CV_BGR2GRAY)
storage = opencv.cvCreateMemStorage(0)
opencv.cvClearMemStorage(storage)
opencv.cvEqualizeHist(grayscale, grayscale)
try:
cascade = opencv.cvLoadHaarClassifierCascade(os.path.join(os.path.dirname(__file__), classifier+".xml"),opencv.cvSize(1,1))
except:
raise AttributeError, "could not load classifier file"
objs = opencv.cvHaarDetectObjects(grayscale, cascade, storage, 1.2, 2, opencv.CV_HAAR_DO_CANNY_PRUNING, opencv.cvSize(50,50))
objects = []
for obj in objs:
objects.append(Haarobj(obj))
opencv.cvReleaseImage(srcimage)
opencv.cvReleaseImage(grayscale)
opencv.cvReleaseMemStorage(storage)
return objects
def detectHaar(iplimage, classifier):
srcimage = opencv.cvCloneImage(iplimage)
grayscale = opencv.cvCreateImage(opencv.cvGetSize(srcimage), 8, 1)
opencv.cvCvtColor(srcimage, grayscale, opencv.CV_BGR2GRAY)
storage = opencv.cvCreateMemStorage(0)
opencv.cvClearMemStorage(storage)
opencv.cvEqualizeHist(grayscale, grayscale)
try:
cascade = opencv.cvLoadHaarClassifierCascade(os.path.join(os.path.dirname(__file__), classifier + ".xml"), opencv.cvSize(1, 1))
except:
raise AttributeError("could not load classifier file")
objs = opencv.cvHaarDetectObjects(grayscale, cascade, storage, 1.2, 2, opencv.CV_HAAR_DO_CANNY_PRUNING, opencv.cvSize(50, 50))
objects = []
for obj in objs:
objects.append(Haarobj(obj))
opencv.cvReleaseImage(srcimage)
opencv.cvReleaseImage(grayscale)
opencv.cvReleaseMemStorage(storage)
return objects
def clone_image(image):
return cv.cvCloneImage(image)
elif cfg.device == 'PR2bag':
''' Instructions: $ roslaunch pr2_lcs_helper get_server2_client.launch
<Opens up services which will be called>
$ roslaunch pr2_lcs_helper laser_assembler_optical_frame_launch.launch
<listens to tilt_scan, provides 'assemble_scan' service, in optical frame
(which is actually left eye)>
$ rosbag play <bag with tf, camera, camera_info, tilt_scan> --clock
$ rosrun laser_camera_segmentation test_display_pr2_try1.py
'''
roslib.load_manifest('pr2_lcs_helper')
import test_uv as PR2_access #rename soon
cloud, roscv_img, pc.pts3d, pc.intensities, camPts, colors, idx_list = \
PR2_access.fetch_PR2_data('all')
N = len(idx_list)
Nbound = len(idx_list[idx_list])
pc.img = cv.cvCloneImage(np.array(roscv_img))
pc.laserscans = []
pc.scan_indices = np.array(range(N))
camPts_bound = camPts[:, idx_list]
pts3d_bound = pc.pts3d[:, idx_list]
scan_indices_bound = np.array(range(N))[idx_list]
intensities_bound = pc.intensities[idx_list]
pc.map = (camPts_bound, camPts, idx_list, pts3d_bound, scan_indices_bound,
intensities_bound)
elif cfg.device == 'PR2':
rospy.init_node('rosSegmentCloud')
print 'STARTING NEW COLLECTION NODE'
ac = acquire_pr2_data.AcquireCloud()
ac.print_test()
print 'FINISHED COLLECTING, SENDING DATA BACK'
def reDraw(self):
pyc.background(255)
pyc.lightSpecular(255 * 30 / 640, 255 * 30 / 640, 255 * 30 / 640)
pyc.directionalLight(255, 255, 255, 0, 0, 1)
pyc.specular(102, 102, 102)
# Project self.pts into image, and display it
imgTmp = cv.cvCloneImage(self.img)
imNP = cv.adaptors.Ipl2NumPy(scale(imgTmp, self.zoom))
color_list = [(255, 255, 0), (255, 0, 0), (0, 255, 255), (0, 255, 0),
(0, 0, 255), (0, 100, 100), (100, 100, 0), (100, 0, 100),
(100, 200, 100), (200, 100, 100), (100, 100, 200),
(100, 0, 200), (0, 200, 100), (0, 100, 200),
(200, 0, 100), (100, 0, 100), (255, 152, 7)]
XformPts = tr.transform(self.transFunc(self.vals), self.pts)
camPts = self.cam_proj_mat * tr.xyzToHomogenous(XformPts)
camPts = camPts / camPts[2]
camPts[0] = (camPts[0] + self.cam_centers[0]) * self.zoom
camPts[1] = (camPts[1] + self.cam_centers[1]) * self.zoom
camPts = np.matrix(np.round(camPts), 'int')
conditions = np.concatenate([
camPts[0] >= 0, camPts[0] < imNP.shape[1], camPts[1] >= 0,
camPts[1] < imNP.shape[0]
], 0)
r, c = np.where(np.all(conditions, 0))
camPts_bound = camPts[:, c.A[0]]
x = np.asarray(self.pts[0])[0][c.A[0]]
x = x - x.min()
x = x / x.max() * 256 #512 #number of colors
x = np.floor(x)
x = np.asarray(np.matrix(x, 'int'))[0]
if self.display_laser:
map2d = np.asarray(camPts_bound[0:2])
n, m = map2d.shape
for i in range(0, m):
imNP[map2d[1, i],
map2d[0, i], :] = [x[i], 256 - x[i],
128 + x[i] / 2] #color_list[x[i]]
imgTmp = cv.adaptors.NumPy2Ipl(imNP)
#imgSmall = cv.cvCreateImage(cv.cvSize(imgTmp.width/3, imgTmp.height/3), cv.IPL_DEPTH_8U, 3)
#cv.cvResize(imgTmp, imgSmall, cv.CV_INTER_AREA)
im = cv.adaptors.Ipl2PIL(imgTmp)
#pyc.rotate(math.radians(90))
pyc.image(im, 0, 0, self.width, self.height)
#pyc.rotate(math.radians(-90))
# Display the current values of the parameter vector (and highlight the selected one)
pyc.textSize(10)
for i, val in enumerate(self.vals):
if np.nonzero(self.selected)[0] == i:
print 'x',
print '%8.4f ' % self.vals[i],
pval = '%7s: %8.4f' % (self.names[i], self.vals[i])
pyc.text(pval, self.width + 15, 20 + 20 * i, 0)
if np.nonzero(self.selected)[0] == i:
pyc.fill(255, 0, 0)
pyc.quad(self.width + 4.0, 15.0 + 20.0 * i - 7.0,
self.width + 4.0, 15.0 + 20.0 * i + 7.0,
self.width + 13.0, 15.0 + 20.0 * i, self.width + 13.0,
15.0 + 20.0 * i)
print '\n'
self.move(pyc.escape_handler(pyc.draw()))
def dilate(cv_image, n_times=1):
dst = cv.cvCloneImage(cv_image)
cv.cvDilate(cv_img, dst, None, n_times)
return dst
#ac.print_test()
print 'FINISHED COLLECTING, SENDING DATA BACK'
np_img, pc.pts3d, pc.intensities, camPts, colors, idx_list = ac.return_data_for_segmentation()
# >> return self.img, self.pts3d, self.intensities,
# self.camPts, self.colors, self.camPts_idx
pc.pts3d = np.array(pc.pts3d)
N = len(idx_list)
Nbound = len( idx_list[idx_list] )
print 'Camera view contains', Nbound, 'of', N, '3D points within its field of view.\n'
if Nbound == 0:
print 'WARNING: Cannot start segmentation if no 3D points overlap image. Exiting.'
print 'This can happen if the pointcloud is empty [], if the transformation is wrong, \
or the camera and scanner are not pointed in the same direction'
import sys; sys.exit()
pc.img = cv.cvCloneImage(np.array(np_img) ) #only works in CV 2.0, but not cv 2.1
pc.laserscans = []
pc.scan_indices = np.array( range(N) )
pc.camPts = camPts
pc.camPts_bound = camPts[:,idx_list]
pc.pts3d_bound = pc.pts3d[:,idx_list]
pc.scan_indices_bound = np.array(range(N))[idx_list]
pc.intensities_bound = pc.intensities[idx_list]
pc.idx_list = idx_list
###pc.map = (pc.camPts_bound, pc.camPts, pc.idx_list, pc.pts3d_bound,pc.scan_indices_bound, pc.intensities_bound)
#create dummy surface polygon to define ROI
# [Below] These polygons will be used to generate a 'mask' cv image and used to
# limit the area of image where 3D projected points are considered.
image_size = cv.cvGetSize(pc.img)
w = image_size.width
def reDraw(self):
pyc.background(255)
pyc.lightSpecular(255*30/640, 255*30/640, 255*30/640)
pyc.directionalLight(255,255,255,0,0,1)
pyc.specular(102, 102, 102)
# Project self.pts into image, and display it
imgTmp = cv.cvCloneImage(self.img)
imNP = cv.adaptors.Ipl2NumPy(scale(imgTmp, self.zoom))
color_list = [(255,255,0),(255,0,0),(0,255,255),(0,255,0),(0,0,255),(0,100,100),(100,100,0),
(100,0,100),(100,200,100),(200,100,100),(100,100,200),(100,0,200),(0,200,100),
(0,100,200),(200,0,100),(100,0,100),(255,152,7) ]
XformPts = tr.transform( self.transFunc(self.vals), self.pts )
camPts = self.cam_proj_mat * tr.xyzToHomogenous(XformPts)
camPts = camPts / camPts[2]
camPts[0] = (camPts[0] + self.cam_centers[0]) * self.zoom
camPts[1] = (camPts[1] + self.cam_centers[1]) * self.zoom
camPts = np.matrix( np.round(camPts), 'int')
conditions = np.concatenate([camPts[0] >= 0,
camPts[0] < imNP.shape[1],
camPts[1] >= 0,
camPts[1] < imNP.shape[0]], 0)
r, c = np.where(np.all(conditions, 0))
camPts_bound = camPts[:, c.A[0]]
x = np.asarray(self.pts[0])[0][c.A[0]]
x = x - x.min()
x = x / x.max() * 256 #512 #number of colors
x = np.floor(x)
x = np.asarray(np.matrix(x,'int'))[0]
if self.display_laser:
map2d = np.asarray(camPts_bound[0:2])
n,m = map2d.shape
for i in range(0,m):
imNP[map2d[1,i],map2d[0,i], :] = [x[i],256-x[i],128+x[i]/2]#color_list[x[i]]
imgTmp = cv.adaptors.NumPy2Ipl(imNP)
#imgSmall = cv.cvCreateImage(cv.cvSize(imgTmp.width/3, imgTmp.height/3), cv.IPL_DEPTH_8U, 3)
#cv.cvResize(imgTmp, imgSmall, cv.CV_INTER_AREA)
im = cv.adaptors.Ipl2PIL(imgTmp)
#pyc.rotate(math.radians(90))
pyc.image(im, 0,0, self.width, self.height)
#pyc.rotate(math.radians(-90))
# Display the current values of the parameter vector (and highlight the selected one)
pyc.textSize(10)
for i, val in enumerate(self.vals):
if np.nonzero(self.selected)[0] == i:
print 'x',
print '%8.4f ' % self.vals[i],
pval = '%7s: %8.4f' % (self.names[i], self.vals[i])
pyc.text(pval, self.width + 15, 20 + 20*i, 0)
if np.nonzero(self.selected)[0] == i:
pyc.fill(255,0,0)
pyc.quad(self.width+4.0, 15.0 + 20.0*i - 7.0,
self.width+4.0, 15.0 + 20.0*i + 7.0,
self.width+13.0, 15.0 + 20.0*i,
self.width+13.0, 15.0 + 20.0*i)
print '\n'
self.move(pyc.escape_handler(pyc.draw()))
'''
if cfg.device=='PR2':
rospy.init_node('rosSegmentCloud')
print 'STARTING NEW COLLECTION NODE'
ac = acquire_pr2_data.AcquireCloud()
#ac.print_test()
print 'FINISHED COLLECTING, SENDING DATA BACK'
np_img, pc.pts3d, pc.intensities, camPts, colors, idx_list = ac.return_data_for_segmentation()
# >> return self.img, self.pts3d, self.intensities,
# self.camPts, self.colors, self.camPts_idx
pc.pts3d = np.array(pc.pts3d)
N = len(idx_list)
Nbound = len( idx_list[idx_list] )
pc.img = cv.cvCloneImage(np.array(np_img) )
pc.laserscans = []
pc.scan_indices = np.array( range(N) )
camPts_bound = camPts[:,idx_list]
pts3d_bound = pc.pts3d[:,idx_list]
scan_indices_bound = np.array(range(N))[idx_list]
intensities_bound = pc.intensities[idx_list]
pc.map = (camPts_bound, camPts, idx_list, pts3d_bound,
scan_indices_bound, intensities_bound)
#create dummy surface polygon to define ROI
# [Below] These polygons will be used to generate a 'mask' cv image and used to
# limit the area of image where 3D projected points are considered.
image_size = cv.cvGetSize(pc.img)
w = image_size.width
h = image_size.height
def prepare(self,
features_k_nearest_neighbors,
nonzero_indices=None,
all_save_load=False,
regenerate_neightborhood_indices=False):
#print np.shape(self.processor.pts3d_bound), 'shape pts3d_bound'
imgTmp = cv.cvCloneImage(self.processor.img)
self.imNP = ut.cv2np(imgTmp, format='BGR')
###self.processor.map2d = np.asarray(self.processor.camPts_bound) #copied from laser to image mapping
if features_k_nearest_neighbors == None or features_k_nearest_neighbors == False: #use range
self.kdtree2d = kdtree.KDTree(self.processor.pts3d_bound.T)
#print len(nonzero_indices)
#print np.shape(np.asarray((self.processor.pts3d_bound.T)[nonzero_indices]))
if nonzero_indices != None:
print ut.getTime(), 'query ball tree for ', len(
nonzero_indices), 'points'
kdtree_query = kdtree.KDTree(
(self.processor.pts3d_bound.T)[nonzero_indices])
else:
print ut.getTime(), 'query ball tree'
kdtree_query = kdtree.KDTree(self.processor.pts3d_bound.T)
filename = self.processor.config.path + '/data/' + self.processor.scan_dataset.id + '_sphere_neighborhood_indices_' + str(
self.processor.feature_radius) + '.pkl'
if all_save_load == True and os.path.exists(
filename) and regenerate_neightborhood_indices == False:
#if its already there, load it:
print ut.getTime(), 'loading', filename
self.kdtree_queried_indices = ut.load_pickle(filename)
else:
self.kdtree_queried_indices = kdtree_query.query_ball_tree(
self.kdtree2d, self.processor.feature_radius, 2.0,
0.2) #approximate
print ut.getTime(), 'queried kdtree: ', len(
self.kdtree_queried_indices
), 'points, radius:', self.processor.feature_radius
if all_save_load == True:
ut.save_pickle(self.kdtree_queried_indices, filename)
#make dict out of list for faster operations? (doesn't seem to change speed significantly):
#self.kdtree_queried_indices = dict(zip(xrange(len(self.kdtree_queried_indices)), self.kdtree_queried_indices))
else: #experiemental: use_20_nearest_neighbors == True
#TODO: exclude invalid values in get_featurevector (uncomment code there)
self.kdtree2d = kdtree.KDTree(self.processor.pts3d_bound.T)
self.kdtree_queried_indices = []
print ut.getTime(
), 'kdtree single queries for kNN start, k=', features_k_nearest_neighbors
count = 0
for point in ((self.processor.pts3d_bound.T)[nonzero_indices]):
count = count + 1
result = self.kdtree2d.query(point,
features_k_nearest_neighbors, 0.2,
2, self.processor.feature_radius)
#existing = result[0][0] != np.Inf
#print existing
#print result[1]
self.kdtree_queried_indices += [result[1]] #[existing]
if count % 4096 == 0:
print ut.getTime(), count
print ut.getTime(), 'kdtree singe queries end'
#convert to numpy array -> faster access
self.kdtree_queried_indices = np.asarray(
self.kdtree_queried_indices)
#print self.kdtree_queried_indices
#takes long to compute:
#avg_len = 0
#minlen = 999999
#maxlen = 0
#for x in self.kdtree_queried_indices:
# avg_len += len(x)
# minlen = min(minlen, len(x))
# maxlen = max(maxlen, len(x))
#avg_len = avg_len / len(self.kdtree_queried_indices)
#print ut.getTime(), "range neighbors: avg_len", avg_len, 'minlen', minlen, 'maxlen', maxlen
#create HSV numpy images:
# compute the hsv version of the image
image_size = cv.cvGetSize(self.processor.img)
img_h = cv.cvCreateImage(image_size, 8, 1)
img_s = cv.cvCreateImage(image_size, 8, 1)
img_v = cv.cvCreateImage(image_size, 8, 1)
img_hsv = cv.cvCreateImage(image_size, 8, 3)
cv.cvCvtColor(self.processor.img, img_hsv, cv.CV_BGR2HSV)
cv.cvSplit(img_hsv, img_h, img_s, img_v, None)
self.imNP_h = ut.cv2np(img_h)
self.imNP_s = ut.cv2np(img_s)
self.imNP_v = ut.cv2np(img_v)
textures = texture_features.eigen_texture(self.processor.img)
self.imNP_tex1 = textures[:, :, 0]
self.imNP_tex2 = textures[:, :, 1]
self.debug_before_first_featurevector = True
self.generate_voi_histogram(self.processor.point_of_interest,
self.processor.voi_width)
def convert_cvimage_to_ROS(image):
imgTmp = cv.cvCloneImage(image)
imNP = ut.cv2np(imgTmp,format='BGR')
ROS_image = np.reshape(imNP,(1,-1))[0]
return ROS_image
def morpho_open(cv_image, n_times=1):
dst = cv.cvCloneImage(cv_image)
dst2 = cv.cvCloneImage(cv_image)
cv.cvErode(cv_image, dst, None, n_times)
cv.cvDilate(dst, dst2, None, n_times)
return dst2
def clone_image(image):
return cv.cvCloneImage(image)
def erode(cv_image, n_times=1):
dst = cv.cvCloneImage(cv_image)
cv.cvErode(cv_img, dst, None, n_times)
return dst
def dilate(cv_image, n_times=1):
dst = cv.cvCloneImage(cv_image)
cv.cvDilate(cv_img, dst, None, n_times)
return dst
def morpho_open(cv_image, n_times=1):
dst = cv.cvCloneImage(cv_image)
dst2 = cv.cvCloneImage(cv_image)
cv.cvErode(cv_image, dst, None, n_times)
cv.cvDilate(dst, dst2, None, n_times)
return dst2
def erode(cv_image, n_times=1):
dst = cv.cvCloneImage(cv_image)
cv.cvErode(cv_img, dst, None, n_times)
return dst
def prepare(self, features_k_nearest_neighbors, nonzero_indices = None, all_save_load = False, regenerate_neightborhood_indices = False):
#print np.shape(self.processor.pts3d_bound), 'shape pts3d_bound'
imgTmp = cv.cvCloneImage(self.processor.img)
self.imNP = ut.cv2np(imgTmp,format='BGR')
###self.processor.map2d = np.asarray(self.processor.camPts_bound) #copied from laser to image mapping
if features_k_nearest_neighbors == None or features_k_nearest_neighbors == False: #use range
self.kdtree2d = kdtree.KDTree(self.processor.pts3d_bound.T)
#print len(nonzero_indices)
#print np.shape(np.asarray((self.processor.pts3d_bound.T)[nonzero_indices]))
if nonzero_indices != None:
print ut.getTime(), 'query ball tree for ', len(nonzero_indices), 'points'
kdtree_query = kdtree.KDTree((self.processor.pts3d_bound.T)[nonzero_indices])
else:
print ut.getTime(), 'query ball tree'
kdtree_query = kdtree.KDTree(self.processor.pts3d_bound.T)
filename = self.processor.config.path+'/data/'+self.processor.scan_dataset.id+'_sphere_neighborhood_indices_'+str(self.processor.feature_radius)+'.pkl'
if all_save_load == True and os.path.exists(filename) and regenerate_neightborhood_indices == False:
#if its already there, load it:
print ut.getTime(), 'loading',filename
self.kdtree_queried_indices = ut.load_pickle(filename)
else:
self.kdtree_queried_indices = kdtree_query.query_ball_tree(self.kdtree2d, self.processor.feature_radius, 2.0, 0.2) #approximate
print ut.getTime(), 'queried kdtree: ',len(self.kdtree_queried_indices),'points, radius:',self.processor.feature_radius
if all_save_load == True:
ut.save_pickle(self.kdtree_queried_indices, filename)
#make dict out of list for faster operations? (doesn't seem to change speed significantly):
#self.kdtree_queried_indices = dict(zip(xrange(len(self.kdtree_queried_indices)), self.kdtree_queried_indices))
else: #experiemental: use_20_nearest_neighbors == True
#TODO: exclude invalid values in get_featurevector (uncomment code there)
self.kdtree2d = kdtree.KDTree(self.processor.pts3d_bound.T)
self.kdtree_queried_indices = []
print ut.getTime(), 'kdtree single queries for kNN start, k=', features_k_nearest_neighbors
count = 0
for point in ((self.processor.pts3d_bound.T)[nonzero_indices]):
count = count + 1
result = self.kdtree2d.query(point, features_k_nearest_neighbors,0.2,2,self.processor.feature_radius)
#existing = result[0][0] != np.Inf
#print existing
#print result[1]
self.kdtree_queried_indices += [result[1]] #[existing]
if count % 4096 == 0:
print ut.getTime(),count
print ut.getTime(), 'kdtree singe queries end'
#convert to numpy array -> faster access
self.kdtree_queried_indices = np.asarray(self.kdtree_queried_indices)
#print self.kdtree_queried_indices
#takes long to compute:
#avg_len = 0
#minlen = 999999
#maxlen = 0
#for x in self.kdtree_queried_indices:
# avg_len += len(x)
# minlen = min(minlen, len(x))
# maxlen = max(maxlen, len(x))
#avg_len = avg_len / len(self.kdtree_queried_indices)
#print ut.getTime(), "range neighbors: avg_len", avg_len, 'minlen', minlen, 'maxlen', maxlen
#create HSV numpy images:
# compute the hsv version of the image
image_size = cv.cvGetSize(self.processor.img)
img_h = cv.cvCreateImage (image_size, 8, 1)
img_s = cv.cvCreateImage (image_size, 8, 1)
img_v = cv.cvCreateImage (image_size, 8, 1)
img_hsv = cv.cvCreateImage (image_size, 8, 3)
cv.cvCvtColor (self.processor.img, img_hsv, cv.CV_BGR2HSV)
cv.cvSplit (img_hsv, img_h, img_s, img_v, None)
self.imNP_h = ut.cv2np(img_h)
self.imNP_s = ut.cv2np(img_s)
self.imNP_v = ut.cv2np(img_v)
textures = texture_features.eigen_texture(self.processor.img)
self.imNP_tex1 = textures[:,:,0]
self.imNP_tex2 = textures[:,:,1]
self.debug_before_first_featurevector = True
self.generate_voi_histogram(self.processor.point_of_interest,self.processor.voi_width)
def jitter(self, i_image, i_n_examples):
"""1) Apply various random affine transform to i_image (scale, rotation),
where i_n_examples is the number of transformations. The affine transforms happen
around the center of the original region of interest.
2) Translate roi with various values - crop the image from this region.
3) The same normalisation is then applied to all the cropped images, specified by setParams"""
#Store transforms applied in format tx ty scale angle
o_transforms = numpy.array([0., 0., 1., 0.])
if self.__roi == None:
print "No region of interest - returning input image!"
return None
#Always return the normalised verion of the input image
image = cv.Ipl2NumPy(self.normalise(i_image))
o_data = numpy.tile(None, (image.shape[0], image.shape[1], 1))
o_data[:, :, 0] = image
if i_n_examples == 0:
return (o_data, o_transforms)
#Rotation point should be around original roi center
center = cv.cvPoint2D32f(self.__roi.x + self.__roi.width / 2,
self.__roi.y + self.__roi.height / 2)
angles = numpy.random.uniform(-30., 30., i_n_examples)
scales = numpy.random.uniform(0.9, 1.1, i_n_examples)
tx = numpy.int32(
numpy.round(numpy.random.uniform(-20, 20, i_n_examples)))
ty = numpy.int32(
numpy.round(numpy.random.uniform(-20, 20, i_n_examples)))
x = self.__roi.x + tx
y = self.__roi.y + ty
#FIXME: Extend valid indices to outliers due to affine transform!!!
min_x = 0
min_y = 0
max_x = i_image.width - self.__roi.width - 1
max_y = i_image.height - self.__roi.height - 1
valid_x = numpy.hstack(
[numpy.nonzero(x >= min_x)[0],
numpy.nonzero(x < max_x)[0]])
valid_y = numpy.hstack(
[numpy.nonzero(y >= min_y)[0],
numpy.nonzero(y < max_y)[0]])
valid_idx = numpy.unique(numpy.hstack([valid_x, valid_y]))
original_roi = cv.cvRect(self.__roi.x, self.__roi.y, self.__roi.width,
self.__roi.height)
if self.__rot_mat == None:
original_rot_matrix = None
else:
original_rot_matrix = cv.cvCloneImage(self.__rot_mat)
for index in valid_idx:
params = numpy.array(
[tx[index], ty[index], scales[index], angles[index]])
self.setAffineTransform(center, scales[index], angles[index])
self.__roi.x = int(x[index])
self.__roi.y = int(y[index])
image = cv.Ipl2NumPy(self.normalise(i_image))
o_data = numpy.dstack([o_data, image])
o_transforms = numpy.vstack([o_transforms, params])
#Restore the original region of interest
self.__roi.x = original_roi.x
self.__roi.y = original_roi.y
if original_rot_matrix == None:
self.__rot_mat = None
else:
self.__rot_mat = cv.cvCloneImage(original_rot_matrix)
return (o_data, o_transforms)
