def get_Npixels_changed(reffile_pathname,compfile_pathname,changethresh):
reffilename,refnameprefix,refseqnumber = celegansmetadata.getjpg_metadata(reffile_pathname)
compfilename,compnameprefix,compseqnumber = celegansmetadata.getjpg_metadata(compfile_pathname)
refimage = get_image(reffile_pathname)
compimage = get_image(compfile_pathname)
pixchange = (refimage - compimage) >= changethresh
Npixels_changed = numpy.count_nonzero(pixchange)
return (refseqnumber,compseqnumber,Npixels_changed)
def do_pixel_change_on_batch(RefAndCompare):
reffilename,refnameprefix,refseqnumber = celegansmetadata.getjpg_metadata(RefAndCompare[0][0])
nametag = str(refseqnumber)
for thresh in ChangeThreshes:
ScratchFile = open(SCRATCHDIR+"/" + nametag + ".thresh" + str(thresh) + ".pixtemp",'w')
for item in RefAndCompare:
refseqnumber,compseqnumber,Npixels_changed = get_Npixels_changed(
item[0],item[1],thresh)
ScratchFile.write("%8i,%8i,%8i \n" % (refseqnumber,compseqnumber,Npixels_changed ))
def process_frame(file_pathname,max_image,THRESH,THRESH_DIFF,NPOINTS,ANGLE_STEP_SIZE,SCRATCHDIR,DOVIZ,VIZDIR,DETMODE,WORMWIDTH,WORMLENGTH):
IP = Store()
IP.scratchdir = SCRATCHDIR
# get the file meta data
IP.file_pathname = file_pathname
IP.FileName,IP.NamePrefix,IP.SeqNumber = celegansmetadata.getjpg_metadata(IP.file_pathname)
# read the image file
IP.OrigImage = libcelegans.get_image(file_pathname)
# perform the image segmentation
IP.DiffImage,IP.WormBinArr,IP.WormArr = segment.doseg(max_image,IP.OrigImage,THRESH,THRESH_DIFF)
# get the 1 pixel wide border of the worm
#'''
IP.OrigBorder = libcelegans.get_border(IP.WormBinArr)
# find the lowest-cost path route around the worm
IP.BorderRoute = libcelegans.get_border_route(IP.OrigBorder)
# reduce the number of points in the path
IP.SimpleBorderRoute = libcelegans.get_simple_border_route(IP.BorderRoute,NPOINTS)
# get a head and tail assignment (by sharpness)
IP.Head,IP.Tail,IP.head,IP.tail = libcelegans.get_border_endpoint(
IP.FileName,IP.OrigBorder,IP.WormBinArr,
IP.SimpleBorderRoute,step_size=ANGLE_STEP_SIZE)
# run the intensity test
IP.PassIntensityTest = libcelegans.perform_head_tail_intensity_test(IP.Head,IP.Tail,IP.WormArr)
# run the volume test
IP.PassVolumeTest = libcelegans.perform_head_tail_volume_test(IP.Head,IP.Tail,IP.WormArr)
# head and tail in hand, use a path path finding algorithm to find the
# lowest-cost path from head to tail, through the border of the worm
IP.SideOnePath,IP.SideTwoPath,IP.SideOneArr,IP.SideTwoArr = \
libcelegans.get_side_paths(IP.Head,IP.Tail,IP.BorderRoute)
# reduce number of points
IP.SimpleSideOnePath = libcelegans.get_simple_side_route(IP.SideOnePath,Nworm_divisions=NPOINTS)
# reduce number of points
IP.SimpleSideTwoPath = libcelegans.get_simple_side_route(IP.SideTwoPath,Nworm_divisions=NPOINTS)
# compute midline
IP.MidLine = libcelegans.get_midline(IP.SimpleSideOnePath,IP.SimpleSideTwoPath)
# test to see if the sides and midline are reasonable
IP.Is_Loop = libcelegans.perform_loop_test(IP.SideOnePath,IP.SideTwoPath)#'''
# write the descriptor file
DescriptorFile = open(SCRATCHDIR+"/"+IP.NamePrefix+".pickle",'wb')
output.serialize_output(IP.SideOnePath,IP.SideTwoPath,IP.SimpleSideOnePath,IP.SimpleSideTwoPath,
IP.MidLine,IP.head,IP.tail,IP.PassIntensityTest,
IP.PassVolumeTest,IP.Is_Loop,DescriptorFile)
# do the visualization
if DOVIZ == 'yes':
visualization.viz(VIZDIR,IP.NamePrefix,IP.DiffImage,IP.OrigImage,IP.SideOnePath,
IP.SideTwoPath,IP.Tail,IP.Head,IP.MidLine)
def process_frame(file_pathname,max_image,THRESH,THRESH_DIFF,NPOINTS,ANGLE_STEP_SIZE,SCRATCHDIR,DOVIZ,VIZDIR,DETMETHOD,WORMWIDTH,WORMLENGTH):
IP = Store()
IP.scratchdir = SCRATCHDIR
# get the file meta data
IP.file_pathname = file_pathname
IP.FileName,IP.NamePrefix,IP.SeqNumber = celegansmetadata.getjpg_metadata(IP.file_pathname)
# read the image file
IP.OrigImage = libcelegans.get_image(file_pathname)
# perform the image segmentation
IP.DiffImage,IP.WormBinArr,IP.WormArr = segment.doseg(max_image,IP.OrigImage,THRESH,THRESH_DIFF)
if DETMETHOD == 0 or DETMETHOD == 1:
# get the 1 pixel wide border of the worm
IP.OrigBorder = libcelegans.get_border(IP.WormBinArr)
# find the lowest-cost path route around the worm
IP.BorderRoute = libcelegans.get_border_route(IP.OrigBorder)
# reduce the number of points in the path
IP.SimpleBorderRoute = libcelegans.get_simple_border_route(IP.BorderRoute,NPOINTS)
# get a head and tail assignment (by sharpness)
IP.Head,IP.Tail,IP.head,IP.tail = libcelegans.get_border_endpoint(
IP.FileName,IP.OrigBorder,IP.WormBinArr,
IP.SimpleBorderRoute,step_size=ANGLE_STEP_SIZE)
# run the intensity test
IP.PassIntensityTest = libcelegans.perform_head_tail_intensity_test(IP.Head,IP.Tail,IP.WormArr)
# run the volume test
IP.PassVolumeTest = libcelegans.perform_head_tail_volume_test(IP.Head,IP.Tail,IP.WormArr)
# head and tail in hand, use a path path finding algorithm to find the
# lowest-cost path from head to tail, through the border of the worm
IP.SideOnePath,IP.SideTwoPath,IP.SideOneArr,IP.SideTwoArr = \
libcelegans.get_side_paths(IP.Head,IP.Tail,IP.BorderRoute)
# reduce number of points
IP.SimpleSideOnePath = libcelegans.get_simple_side_route(IP.SideOnePath,Nworm_divisions=NPOINTS)
# reduce number of points
IP.SimpleSideTwoPath = libcelegans.get_simple_side_route(IP.SideTwoPath,Nworm_divisions=NPOINTS)
# compute midline
IP.MidLine = libcelegans.get_midline(IP.SimpleSideOnePath,IP.SimpleSideTwoPath)
# test to see if the sides and midline are reasonable
IP.Is_Loop = libcelegans.perform_loop_test(IP.SideOnePath,IP.SideTwoPath)
# do the visualization
if DOVIZ == 'yes':
visualization.viz(VIZDIR,IP.NamePrefix,IP.DiffImage,IP.OrigImage,IP.SideOnePath,
IP.SideTwoPath,IP.Tail,IP.Head,IP.MidLine)
if DETMETHOD == 0 or DETMETHOD == 2:
# begin a section for Marc's development code. Pass the IP object with all properties that have already been
# computed for the image.
IP.MarcSides,IP.MarcMidLine,IP.MarcOffsets,IP.MarcScore = marcdev.main(IP,max_image,WORMWIDTH,WORMLENGTH)
# convert MarcMidLine from the TrimImage coordinate system to the OrigImage coordinate system
IP.OffsetMarcMidLine = [(pt[0]+IP.MarcOffsets['ystart'],pt[1]+IP.MarcOffsets['xstart']) for pt in IP.MarcMidLine]
IP.OffsetMarcSides = [(pt[0]+IP.MarcOffsets['ystart'],pt[1]+IP.MarcOffsets['xstart']) for pt in IP.MarcSides]
# do the visualization
if DOVIZ == 'yes':
visualization.mviz(VIZDIR,IP.NamePrefix,IP.OrigImage,IP.OffsetMarcMidLine[-1],IP.OffsetMarcMidLine[0],\
IP.OffsetMarcMidLine,IP.OffsetMarcSides)
# write the descriptor file
DescriptorFile = open(SCRATCHDIR+"/"+IP.NamePrefix+".pickle",'wb')
if DETMETHOD == 0:
output.serialize_output(IP.SideOnePath,IP.SideTwoPath,IP.SimpleSideOnePath,IP.SimpleSideTwoPath,
IP.MidLine,IP.head,IP.tail,IP.PassIntensityTest,IP.PassVolumeTest,IP.Is_Loop,
IP.OffsetMarcMidLine,IP.MarcScore,DescriptorFile)
elif DETMETHOD == 1:
output.serialize_output(IP.SideOnePath,IP.SideTwoPath,IP.SimpleSideOnePath,IP.SimpleSideTwoPath,
IP.MidLine,IP.head,IP.tail,IP.PassIntensityTest,IP.PassVolumeTest,IP.Is_Loop,
None,None,DescriptorFile)
elif DETMETHOD == 2:
output.serialize_output(None,None,None,None,
None,None,None,None,None,None,
IP.OffsetMarcMidLine,IP.MarcScore,DescriptorFile)
def process_frame(file_pathname, max_image, THRESH, THRESH_DIFF, NPOINTS,
ANGLE_STEP_SIZE, SCRATCHDIR, DOVIZ, VIZDIR, DETMODE,
WORMWIDTH, WORMLENGTH):
IP = Store()
IP.scratchdir = SCRATCHDIR
# get the file meta data
IP.file_pathname = file_pathname
IP.FileName, IP.NamePrefix, IP.SeqNumber = celegansmetadata.getjpg_metadata(
IP.file_pathname)
# read the image file
IP.OrigImage = libcelegans.get_image(file_pathname)
# perform the image segmentation
IP.DiffImage, IP.WormBinArr, IP.WormArr = segment.doseg(
max_image, IP.OrigImage, THRESH, THRESH_DIFF)
# get the 1 pixel wide border of the worm
#'''
IP.OrigBorder = libcelegans.get_border(IP.WormBinArr)
# find the lowest-cost path route around the worm
IP.BorderRoute = libcelegans.get_border_route(IP.OrigBorder)
# reduce the number of points in the path
IP.SimpleBorderRoute = libcelegans.get_simple_border_route(
IP.BorderRoute, NPOINTS)
# get a head and tail assignment (by sharpness)
IP.Head, IP.Tail, IP.head, IP.tail = libcelegans.get_border_endpoint(
IP.FileName,
IP.OrigBorder,
IP.WormBinArr,
IP.SimpleBorderRoute,
step_size=ANGLE_STEP_SIZE)
# run the intensity test
IP.PassIntensityTest = libcelegans.perform_head_tail_intensity_test(
IP.Head, IP.Tail, IP.WormArr)
# run the volume test
IP.PassVolumeTest = libcelegans.perform_head_tail_volume_test(
IP.Head, IP.Tail, IP.WormArr)
# head and tail in hand, use a path path finding algorithm to find the
# lowest-cost path from head to tail, through the border of the worm
IP.SideOnePath,IP.SideTwoPath,IP.SideOneArr,IP.SideTwoArr = \
libcelegans.get_side_paths(IP.Head,IP.Tail,IP.BorderRoute)
# reduce number of points
IP.SimpleSideOnePath = libcelegans.get_simple_side_route(
IP.SideOnePath, Nworm_divisions=NPOINTS)
# reduce number of points
IP.SimpleSideTwoPath = libcelegans.get_simple_side_route(
IP.SideTwoPath, Nworm_divisions=NPOINTS)
# compute midline
IP.MidLine = libcelegans.get_midline(IP.SimpleSideOnePath,
IP.SimpleSideTwoPath)
# test to see if the sides and midline are reasonable
IP.Is_Loop = libcelegans.perform_loop_test(IP.SideOnePath,
IP.SideTwoPath) #'''
# write the descriptor file
DescriptorFile = open(SCRATCHDIR + "/" + IP.NamePrefix + ".pickle", 'wb')
output.serialize_output(IP.SideOnePath, IP.SideTwoPath,
IP.SimpleSideOnePath, IP.SimpleSideTwoPath,
IP.MidLine, IP.head, IP.tail, IP.PassIntensityTest,
IP.PassVolumeTest, IP.Is_Loop, DescriptorFile)
# do the visualization
if DOVIZ == 'yes':
visualization.viz(VIZDIR, IP.NamePrefix, IP.DiffImage, IP.OrigImage,
IP.SideOnePath, IP.SideTwoPath, IP.Tail, IP.Head,
IP.MidLine)
