def run(self):
if self.outFile is not None:
#self.outFile.write("{}\t{}\t{}\t{}\n".format('obj', 'view', '\t'.join(["{}_mean".format(label) for label in self.visSys.featureLabels]), '\t'.join(["{}_sd".format(label) for label in self.visSys.featureLabels]))) # vector mean and SD
self.outFile.write("{}\t{}\t{}\n".format('obj', 'view', '\t'.join(["{}_{}".format(label, i) for label in self.visSys.featureLabels for i in xrange(self.visSys.num_feature_neurons)]))) # matrix mean
if self.context.isDir: # input source is a directory
# * Run visual input using manager, looping over all specified object images
for obj in self.objRange:
for view in self.viewRange:
# ** Build image file path from object ID and view angle
input_file = os.path.join(self.inDir, "{}{}{}{}.{}".format(self.input_file_prefix, obj, self.input_file_sep, view, self.input_file_ext))
#assert os.path.exists(input_file), "Input file \"{}\" doesn't exist".format(input_file)
if not os.path.exists(input_file):
self.logger.warn("Input file \"{}\" doesn't exist".format(input_file))
continue
self.logger.info("Input file: {}".format(input_file))
# ** Modify context to set image file as input source, and run it through the visual system
self.context.options.input_source = input_file
self.context.isImage = True
print "Running..."
run(self.visMan, resetContextTime=False) # use the same manager so that visual system is only created once
if self.outFile is not None:
#self.outFile.write("{}\t{}\t{}\t{}\n".format(obj, view, '\t'.join(str(feat_mean) for feat_mean in self.visMan.featureVectorMean), '\t'.join(str(feat_sd) for feat_sd in self.visMan.featureVectorSD))) # vector mean and SD
self.outFile.write("{}\t{}\t{}\n".format(obj, view, '\t'.join(str(feat) for feat in self.visMan.featureMatrixMean.flat))) # matrix mean
else:
run(self.visMan, resetContextTime=False) # run on the sole input source (image or video)
if self.outFile is not None:
self.outFile.close()
self.logger.info("Output file closed.")
def test_rod_potential(self):
from ..neuron import action_potential_trough, action_potential_peak
class MonitoringProjector(Projector):
do_plot = False
def __init__(self, retina=None):
Projector.__init__(self, retina)
# Neuron to monitor
self.testRodIdx = 0
self.testRod = self.target.rods.neurons[self.testRodIdx]
self.logger.info("Test rod [{}]: {}".format(self.testRodIdx, self.testRod))
# Plotting
if self.do_plot:
self.logger.info("Plotting is enabled")
self.fig = figure()
hold(True)
self.ax = self.fig.gca()
self.ax.set_ylim(action_potential_trough.mu - 0.01, action_potential_peak + 0.02)
self.ax.set_title("Neuron")
self.ax.set_xlabel("Time (s)")
self.ax.set_ylabel("Membrane potential (V)")
def process(self, imageIn, timeNow):
keepRunning, imageOut = Projector.process(self, imageIn, timeNow)
self.testRod.p = 1.0 # make sure it updated every iteration
if self.do_plot:
self.testRod.plot()
pause(0.01)
print "{}\t{}\t{}\t{}\t{}\t{}".format(timeNow, self.testRod.response, self.testRod.potential, self.testRod.I_e, self.testRod.expDecayFactor, self.testRod.pixelValue) # [debug, non-GUI]
#print "{}\t{}\t{}\t{}".format(timeNow, self.testRod.potential, self.testRod.expDecayFactor, self.testRod.pixelValue) # [debug, non-GUI, for BipolarCells]
#cv2.circle(imageOut, (self.testRod.pixel[0], self.testRod.pixel[1]), 3, np.uint8([255, 0, 255]))
imageOut[self.testRod.pixel[1], self.testRod.pixel[0]] = np.uint8([255, 0, 255])
return keepRunning, imageOut
def onKeyPress(self, key, keyChar=None):
if keyChar == '.':
self.testRodIdx = (self.testRodIdx + 1) % len(self.target.rods.neurons)
self.testRod = self.target.rods.neurons[self.testRodIdx]
self.logger.info("[>] Test rod [{}]: {}".format(self.testRodIdx, self.testRod))
elif keyChar == ',':
self.testRodIdx = (self.testRodIdx - 1) % len(self.target.rods.neurons)
self.testRod = self.target.rods.neurons[self.testRodIdx]
self.logger.info("[<] Test rod [{}]: {}".format(self.testRodIdx, self.testRod))
else:
return Projector.onKeyPress(self, key, keyChar)
return True
print "Running MonitoringProjector instance..."
run(MonitoringProjector(Projector), description="Retina processing with monitor on a single neuron.")
# * TODO Compute feature vector of attended region
# * Show output images if in GUI mode
if self.context.options.gui:
#cv2.imshow("Hue", self.images['H'])
#cv2.imshow("Saturation", self.images['S'])
#cv2.imshow("Value", self.images['V'])
cv2.imshow("Rod response", self.imageRod)
cv2.imshow("S-cone response", self.imagesCone['S'])
cv2.imshow("M-cone response", self.imagesCone['M'])
cv2.imshow("L-cone response", self.imagesCone['L'])
cv2.imshow("ON Bipolar cells", self.imagesBipolar['ON'])
cv2.imshow("OFF Bipolar cells", self.imagesBipolar['OFF'])
#cv2.imshow("ON Ganglion cells", self.imagesGanglion['ON'])
#cv2.imshow("OFF Ganglion cells", self.imagesGanglion['OFF'])
for ganglionType, ganglionImage in self.imagesGanglion.iteritems():
cv2.imshow("{} Ganglion cells".format(ganglionType), ganglionImage)
cv2.imshow("Salience", self.imageSalience)
# Designate a representative output image
self.imageOut = self.imageSalience
#_, self.imageOut = cv2.threshold(self.imageOut, 0.15, 1.0, cv2.THRESH_TOZERO) # apply threshold to remove low-response regions
return True, self.imageOut
if __name__ == "__main__":
Context.createInstance(description="Test application that uses a SimplifiedProjector to run image input through a (simplified) Retina.")
run(Projector(Retina()))
import cv2 # OpenCV functions
import cv2.cv as cv # OpenCV constants
from lumos.base import FrameProcessor # base processor class
from lumos.input import run # driver function
class MyAwesomeProcessor(FrameProcessor):
"""Custom processor that selects hues based on current time."""
def process(self, imageIn, timeNow):
# Convert input Red-Green-Blue image to Hue-Saturation-Value
hsv = cv2.cvtColor(imageIn, cv.CV_BGR2HSV)
h, s, v = cv2.split(hsv) # split into 3 channels
h = h.reshape(h.shape + (1, ))
# Pick desired hue range based on current time
hue = int(((timeNow % 10) / 10) * 180)
min_hue = max(0, hue - 10)
max_hue = min(180, hue + 10)
# Apply mask to select pixels in hue range and return
mask = cv2.inRange(h, min_hue, max_hue)
imageOut = cv2.bitwise_and(imageIn, imageIn, mask=mask)
return True, imageOut
if __name__ == "__main__":
# Run a custom processor instance (NOTE pass in class name)
run(MyAwesomeProcessor, description="A sample lumos application")
def test_projector(self): run(Projector(Retina()), description="Test application that uses a Projector to run image input through a Retina.")
import cv2 # OpenCV functions
import cv2.cv as cv # OpenCV constants
from lumos.base import FrameProcessor # base processor class
from lumos.input import run # driver function
class MyAwesomeProcessor(FrameProcessor):
"""Custom processor that selects hues based on current time."""
def process(self, imageIn, timeNow):
# Convert input Red-Green-Blue image to Hue-Saturation-Value
hsv = cv2.cvtColor(imageIn, cv.CV_BGR2HSV)
h, s, v = cv2.split(hsv) # split into 3 channels
h = h.reshape(h.shape + (1,))
# Pick desired hue range based on current time
hue = int(((timeNow % 10) / 10) * 180)
min_hue = max(0, hue - 10)
max_hue = min(180, hue + 10)
# Apply mask to select pixels in hue range and return
mask = cv2.inRange(h, min_hue, max_hue)
imageOut = cv2.bitwise_and(imageIn, imageIn, mask=mask)
return True, imageOut
if __name__ == "__main__":
# Run a custom processor instance (NOTE pass in class name)
run(MyAwesomeProcessor, description="A sample lumos application")
