def readingVideo(videoFile):
Color = preprocessing.ColorDetection()
File = preprocessing.File()
RegionGrowing = preprocessing.RegionGrowing()
ImageProcessing = preprocessing.ImageProcessing()
Moving = preprocessing.Moving()
stdDev, mean = Color.getStdDevAndMean('-dataset-fire_image')
print "Stdev : "+str(stdDev), "Mean : "+str(mean)
print "Video Frame : ",File.getCountFrame(videoFile)
print "Video Size : ",len(File.readVideo(videoFile)[1]),len(File.readVideo(videoFile)[1][0])
classifier = cls.getClassifier('-file-datatraining/TA.xls','rbf',5)
fireFrame = numpy.array([0,0,0,0])
list_wavelet = []
list_color = Color.getFireArray('-dataset-fire_file/color_5x10^-9.txt')
AllFrame = 0
counter = 0
starts = time.time()
while(File.isOpened(videoFile)):
try :
#get curent frame
currentFrame = File.readVideo(videoFile)[1]
if len(currentFrame) == 0:
return
currentFrame2 = copy.copy(currentFrame)
currentFrame = ImageProcessing.getDownSize(currentFrame)
counter+=1
File.saveImage('-code-approving/'+str(counter)+'.png',currentFrame)
File.saveImage('-code-approving/'+str(counter)+'_2.png',currentFrame2)
# step 1 get moving pixel
movingFrame = Moving.getMovingForeGround(copy.copy(currentFrame))
movingPixel = Moving.getMovingCandidatePixel(movingFrame)
mvng = mv.getMovingForeGroundColor(currentFrame,movingFrame)
File.saveImage('-code-approving/mvng'+str(counter)+'.png',mvng)
# step 2 candidate pixel ( color probability )
ColorCandidatePixel = Color.getColorCandidatePixel(copy.copy(movingPixel), copy.copy(currentFrame), list_color)
clr = mv.delPixel(ColorCandidatePixel[1], mvng)
File.saveImage('-code-approving/clr'+str(counter)+'.png',clr)
#region growing
region = RegionGrowing.getRegionGrowing(ColorCandidatePixel[0], copy.copy(currentFrame),list_color,counter)
reg = copy.copy(currentFrame)
for y in range(0,len(region)):
for x in range(0,len(region[y])):
if region[y][x] == 0:
reg[y][x] = [0,0,0]
# step 3 region candidate pixel ( region size )
sizeRegionCandidatePixel = RegionGrowing.getFilterSizeRegion(copy.copy(ColorCandidatePixel[0]),copy.copy(region))
siz = mv.delPixel(sizeRegionCandidatePixel[1], clr)
File.saveImage('-code-approving/siz'+str(counter)+'.png',siz)
#preparing classification
grayImage = ImageProcessing.getRGBtoGray(currentFrame2)
LL,(HL,LH,HH) = wv.toWavelet(copy.copy(grayImage))
list_wavelet.append([HL,LH,HH])
if (counter<=10):
continue
list_wavelet.pop(0)
FinalCandidatePixel = cls.doClassification(classifier,copy.copy(sizeRegionCandidatePixel[0]),list_wavelet)
fireFrameImage = Moving.markingFire(FinalCandidatePixel[0],currentFrame2, 2)
fre = mv.delPixel(FinalCandidatePixel[1], siz)
File.saveImage('-code-approving/fre'+str(counter)+'.png',fre)
# fireFrameImage = Moving.markingFire(FinalCandidatePixel[0],currentFrame)
File.showVideo('Final',fireFrameImage)
File.saveImage('-code-approving/fnl'+str(counter)+'.png',fireFrameImage)
if len(movingPixel[0])>0:
fireFrame[0]+=1
if len(ColorCandidatePixel[0])>0:
fireFrame[1]+=1
if len(sizeRegionCandidatePixel[0])>0:
fireFrame[2]+=1
if len(FinalCandidatePixel[0])>0:
fireFrame[3]+=1
AllFrame+=1
File.waitVideo(1)
except :
print "Time : ",time.time() - starts
return (fireFrame)/float(AllFrame)
print "Time : ",time.time() - starts
return (fireFrame)/float(AllFrame)
# list_variance.append(['-dataset-fire_file/color_5x10^-9.txt',5,'rbf','4_1.xls']) #5%
# list_variance.append(['-dataset-fire_file/color_5x10^-9.txt',5,'rbf','4_2.xls']) #10%
# list_variance.append(['-dataset-fire_file/color_5x10^-9.txt',5,'rbf','5_1.xls']) #240 x 320
# list_variance.append(['-dataset-fire_file/color_5x10^-9.txt',5,'rbf','5_2.xls']) #120 x 160
# list_variance.append(['-dataset-fire_file/color_5x10^-9.txt',5,'rbf','original.xls']) #originalpaper
for variasi in list_variance:
print variasi
path = '../data uji/'
File = preprocessing.File()
Color = preprocessing.ColorDetection()
list_file = File.readFolder(path)
classifier = cls.getClassifier('-file-datatraining/TA.xls',variasi[2],variasi[1])
list_color = Color.getFireArray(variasi[0])
for x in list_file:
preprocessing = reload(preprocessing)
print x
fileName = path+x
videoFile = File.openVideo(fileName)
res, frameCounter, times = readingVideo(videoFile,list_color,classifier,x)
res*=100
print "Acc : ",res[3],' %'
report = []
report.append(x)
for y in res:
report.append(y)
report.append('')
report.append(frameCounter)
def readingVideo(videoFile):
Color = preprocessing.ColorDetection()
File = preprocessing.File()
RegionGrowing = preprocessing.RegionGrowing()
ImageProcessing = preprocessing.ImageProcessing()
Moving = preprocessing.Moving()
stdDev, mean = Color.getStdDevAndMean("-dataset-fire_image")
print "Stdev : " + str(stdDev), "Mean : " + str(mean)
print "Video Frame : ", File.getCountFrame(videoFile)
print "Video Size : ", len(File.readVideo(videoFile)[1]), len(File.readVideo(videoFile)[1][0])
classifier = cls.getClassifier("-file-datatraining/TA.xls", "rbf", 5)
fireFrame = numpy.array([0, 0, 0, 0])
list_wavelet = []
list_color = Color.getFireArray("-dataset-fire_file/color_5x10^-9.txt")
AllFrame = 0
counter = 0
starts = time.time()
while File.isOpened(videoFile):
try:
# get curent frame
currentFrame = File.readVideo(videoFile)[1]
if len(currentFrame) == 0:
return
currentFrame2 = copy.copy(currentFrame)
currentFrame = ImageProcessing.getDownSize(currentFrame)
counter += 1
# step 1 get moving pixel
movingFrame = Moving.getMovingForeGround(copy.copy(currentFrame))
movingPixel = Moving.getMovingCandidatePixel(movingFrame)
# step 2 candidate pixel ( color probability )
ColorCandidatePixel = Color.getColorCandidatePixel(
copy.copy(movingPixel), copy.copy(currentFrame), list_color
)
# region growing
region = RegionGrowing.getRegionGrowing(
ColorCandidatePixel[0], copy.copy(currentFrame), list_color, counter
)
# step 3 region candidate pixel ( region size )
sizeRegionCandidatePixel = RegionGrowing.getFilterSizeRegion(
copy.copy(ColorCandidatePixel[0]), copy.copy(region)
)
# preparing classification
grayImage = ImageProcessing.getRGBtoGray(currentFrame2)
LL, (HL, LH, HH) = wv.toWavelet(copy.copy(grayImage))
list_wavelet.append([HL, LH, HH])
if counter <= 10:
continue
list_wavelet.pop(0)
FinalCandidatePixel = cls.doClassification(classifier, copy.copy(sizeRegionCandidatePixel[0]), list_wavelet)
fireFrameImage = Moving.markingFire(FinalCandidatePixel[0], currentFrame2, 2)
# fireFrameImage = Moving.markingFire2(FinalCandidatePixel[0],currentFrame)
File.showVideo("Final", fireFrameImage)
if len(movingPixel[0]) > 0:
fireFrame[0] += 1
if len(ColorCandidatePixel[0]) > 0:
fireFrame[1] += 1
if len(sizeRegionCandidatePixel[0]) > 0:
fireFrame[2] += 1
if len(FinalCandidatePixel[0]) > 0:
fireFrame[3] += 1
AllFrame += 1
File.waitVideo(1)
except:
print "Time : ", time.time() - starts
return (fireFrame) / float(AllFrame)
print "Time : ", time.time() - starts
return (fireFrame) / float(AllFrame)
def readingVideo(videoFile):
Color = preprocessing.ColorDetection()
File = preprocessing.File()
Intensity = preprocessing.Intensity()
Luminance = preprocessing.Luminance()
RegionGrowing = preprocessing.RegionGrowing()
ImageProcessing = preprocessing.ImageProcessing()
Moving = preprocessing.Moving()
stdDev, mean = Color.getStdDevAndMean('-dataset-fire_image')
print "Stdev : "+str(stdDev), "Mean : "+str(mean)
print "Video Frame : ",File.getCountFrame(videoFile)
print "Video Size : ",len(File.readVideo(videoFile)[1]),len(File.readVideo(videoFile)[1][0])
classifier = cls.getClassifier('-file-datatraining/TA.xls')
fireFrame = numpy.array([0,0,0,0,0,0,0])
list_wavelet = []
list_luminance = []
list_gray_image = []
list_region = []
list_color = Color.getFireArray('color.txt')
AllFrame = 0
counter = 0
starts = time.time()
while(File.isOpened(videoFile)):
try :
#get curent frame
currentFrame = File.readVideo(videoFile)[1]
#compres image
while (len(currentFrame)>150):
if (len(currentFrame)<=300):
currentFrame2 = copy.copy(currentFrame)
currentFrame = ImageProcessing.getDownSize(currentFrame)
counter+=1
gray_image = ImageProcessing.getRGBtoGray(copy.copy(currentFrame))
# step 1 get moving pixel
movingFrame = Moving.getMovingForeGround(copy.copy(currentFrame))
movingPixel = Moving.getMovingCandidatePixel(movingFrame)
# moving = mv.getMovingForeGroundColor(currentFrame,movingFrame)
# File.saveImage('tmp/'+str(counter)+'_.png',moving)
# File.saveImage('tmp/'+str(counter)+'.png',currentFrame)
# step 2 candidate pixel ( color probability )
ColorCandidatePixel = Color.getColorCandidatePixel(copy.copy(movingPixel), copy.copy(currentFrame), list_color)
#region growing
region = RegionGrowing.getRegionGrowing(ColorCandidatePixel[0], copy.copy(currentFrame),list_color,counter)
# step 3 candidate pixel ( brightness ), convert image to gray with luminance and split by region
LuminanceCandidatePixel = ColorCandidatePixel
VarianceCandidatePixel = RegionGrowing.getVarianceColorCandidatePixel(copy.copy(currentFrame),copy.copy(LuminanceCandidatePixel[0]),copy.copy(region))
fireFrameImage = (ImageProcessing.getUpSize(Moving.markingFire(VarianceCandidatePixel[0],currentFrame2, 2)))
File.showVideo('Final',fireFrameImage)
File.waitVideo(1)
except :
return