def write(self, data):
"""
Write the given data to the database.
:type data: dict
:rtype: None
"""
utils.saveJson(self.path(), data)
def main():
src = sys.argv[1]
dest = sys.argv[2]
stats = utils.loadJson(sys.argv[3])
threshold = float(sys.argv[4])
# keep executing only if stats were loaded
if stats != None:
stats = stats['classes']
cache = utils.loadJson(getCacheName(stats))
if cache == None:
cache = {}
applySegmentation(src, dest, threshold, stats, cache)
utils.saveJson(cache, getCacheName(stats))
else:
print('No stats available, cant process images')
print('Finished')
def main():
src = sys.argv[1]
dest = sys.argv[2]
pixels = {}
for f in os.listdir(src):
print('Processing ' + f)
data = utils.loadJson(src + f)
for key in data:
if key != 'height' and key != 'width' and len(data[key]) > 0:
if not key in pixels:
pixels[key] = []
extracted = utils.extractRGB(data[key])
pixels[key] = pixels[key] + extracted
for key in pixels:
print('Extracted ' + str(len(pixels[key])) + ' pixels for ' + key)
print('Saving')
utils.saveJson(pixels, dest + 'pixels.json')
print('Finished')
def main():
src = sys.argv[1]
dest = sys.argv[2]
attempts = 10
ratio = 0.5
thres = 0.95
pixels = utils.loadJson(src)
bestSamples = {}
bestStats = {'classes': []}
minVal = [1e10000]
minTrain = []
minTest = []
for n in range(attempts):
print('** Attempt ' + str(n))
samples = {}
stats = {'classes': []}
# get sets and stats from data
for key in pixels:
print('\tProcessing color ' + key)
sets = getBestSets(key, pixels[key], ratio, attempts)
stats['classes'].append(sets['stats'])
samples[key] = sets
# calculate the classification error
totalTrain = 0.0
totalVal = 0.0
totalTest = 0.0
errTrain = 0
errVal = 0
errTest = 0
cache = {}
for key in samples:
clsTrain = classify(samples[key]['train'], thres, stats['classes'], cache)
errTrain += classificationError(clsTrain, key)
totalTrain += len(samples[key]['train'][utils.R])
clsVal = classify(samples[key]['val'], thres, stats['classes'], cache)
errVal += classificationError(clsVal, key)
totalVal += len(samples[key]['val'][utils.R])
clsTest = classify(samples[key]['test'], thres, stats['classes'], cache)
errTest += classificationError(clsTest, key)
totalTest += len(samples[key]['test'][utils.R])
print('\tError train: \t' + str(errTrain) + '\t' + str(errTrain / totalTrain))
print('\tError val: \t' + str(errVal) + '\t' + str(errVal / totalVal))
print('\tError test: \t' + str(errTest) + '\t' + str(errTest / totalTest))
if errVal < minVal[0]:
minVal = [errVal, totalVal]
minTrain = [errTrain, totalTrain]
minTest = [errTest, totalTest]
bestSamples = samples
bestStats = stats
if minVal[0] == 0:
break
print('Min error')
print('\ttrain: \t' + str(minTrain[0]) + '\t' + str(minTrain[0] / minTrain[1]))
print('\tval: \t' + str(minVal[0]) + '\t' + str(minVal[0] / minVal[1]))
print('\ttest: \t' + str(minTest[0]) + '\t' + str(minTest[0] / minTest[1]))
bestSamples = {'samples': bestSamples, 'errors': {}}
bestSamples['errors']['train'] = minTrain
bestSamples['errors']['val'] = minVal
bestSamples['errors']['test'] = minTest
print('Saving data')
utils.saveJson(bestStats, dest + 'learnedStats.json')
utils.saveJson(bestSamples, dest + 'learnedSamples.json')
print('Finished')
def main():
src = sys.argv[1]
dest = sys.argv[2]
attempts = 10
ratio = 0.5
thres = 0.95
pixels = utils.loadJson(src)
bestSamples = {}
bestStats = {'classes': []}
minVal = [1e10000]
minTrain = []
minTest = []
for n in range(attempts):
print('** Attempt ' + str(n))
samples = {}
stats = {'classes': []}
# get sets and stats from data
for key in pixels:
print('\tProcessing color ' + key)
sets = getBestSets(key, pixels[key], ratio, attempts)
stats['classes'].append(sets['stats'])
samples[key] = sets
# calculate the classification error
totalTrain = 0.0
totalVal = 0.0
totalTest = 0.0
errTrain = 0
errVal = 0
errTest = 0
cache = {}
for key in samples:
clsTrain = classify(samples[key]['train'], thres, stats['classes'],
cache)
errTrain += classificationError(clsTrain, key)
totalTrain += len(samples[key]['train'][utils.R])
clsVal = classify(samples[key]['val'], thres, stats['classes'],
cache)
errVal += classificationError(clsVal, key)
totalVal += len(samples[key]['val'][utils.R])
clsTest = classify(samples[key]['test'], thres, stats['classes'],
cache)
errTest += classificationError(clsTest, key)
totalTest += len(samples[key]['test'][utils.R])
print('\tError train: \t' + str(errTrain) + '\t' +
str(errTrain / totalTrain))
print('\tError val: \t' + str(errVal) + '\t' + str(errVal / totalVal))
print('\tError test: \t' + str(errTest) + '\t' +
str(errTest / totalTest))
if errVal < minVal[0]:
minVal = [errVal, totalVal]
minTrain = [errTrain, totalTrain]
minTest = [errTest, totalTest]
bestSamples = samples
bestStats = stats
if minVal[0] == 0:
break
print('Min error')
print('\ttrain: \t' + str(minTrain[0]) + '\t' +
str(minTrain[0] / minTrain[1]))
print('\tval: \t' + str(minVal[0]) + '\t' + str(minVal[0] / minVal[1]))
print('\ttest: \t' + str(minTest[0]) + '\t' + str(minTest[0] / minTest[1]))
bestSamples = {'samples': bestSamples, 'errors': {}}
bestSamples['errors']['train'] = minTrain
bestSamples['errors']['val'] = minVal
bestSamples['errors']['test'] = minTest
print('Saving data')
utils.saveJson(bestStats, dest + 'learnedStats.json')
utils.saveJson(bestSamples, dest + 'learnedSamples.json')
print('Finished')
# update current usage
book = utils.openGoogleSheet(gSheetId)
dataSheet = book.get_worksheet(0)
date = dataSheet.acell(DATE_CELL).value
cap = int(dataSheet.acell(CAP_CELL).value.split()[0])
target = int(dataSheet.acell(TARGET_CELL).value.split()[0])
usage = int(dataSheet.acell(USAGE_CELL).value.split()[0])
today = int(dataSheet.acell(TODAY_CELL).value.split()[0])
# build data
data = {
'date': date,
'cap': cap,
'usage': usage,
'today': today,
'warning': int(target*GRAPH_WARNING),
'error': target,
'gsheet': gSheetUrl
}
# write cache
utils.saveJson(CACHE_USAGE, data)
except:
data = None
# echo
print('Content-Type: application/json')
print()
print(json.dumps(data))
2.2 Gets the locations in which the image has Faces
2.3 Adds the locations in the global JSON_LIST
3. Dumps the JSON_LIST to results.json
4. STOP
"""
args = parse_args()
TestImgsFolder = args.string
TestImgs = getImages(TestImgsFolder)
for i in range(0, len(TestImgs)):
imgtest = cv2.imread(TestImgsFolder + '/' + TestImgs[i])
gray = cv2.cvtColor(imgtest, cv2.COLOR_RGB2GRAY)
locations = getFaceLocations(gray)
if locations is not None:
locations = nonMaximalSupression(np.array(locations), 0.25)
addToJson(TestImgs[i], locations)
saveJson(JSON_LIST)
