def check(tops_filename, pants_filename, skirts_filename):
file_all_recorder = open('datasets/final-rank/' + account.get_signal() +
'/download/recorder.csv',
'r',
newline='')
recorder_to_read = csv.reader(file_all_recorder)
all_recorder = [row[0] for row in recorder_to_read][1:]
file_all_recorder.close()
file_recorder_tops = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_coat.csv',
'r',
newline='')
recorder_to_read = csv.reader(file_recorder_tops)
tops_recorder = [row[0].split('/')[-1] for row in recorder_to_read]
file_recorder_tops.close()
file_pants_recorder = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_pant.csv',
'r',
newline='')
recorder_to_read = csv.reader(file_pants_recorder)
pants_recorder = [row[0].split('/')[-1] for row in recorder_to_read]
file_pants_recorder.close()
file_skirts_recorder = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_skirt.csv',
'r',
newline='')
recorder_to_read = csv.reader(file_skirts_recorder)
skirts_recorder = [row[0].split('/')[-1] for row in recorder_to_read]
file_skirts_recorder.close()
# 新上传的内容
add_tops = check_add(all_recorder, tops_filename)
add_pants = check_add(all_recorder, pants_filename)
add_skirts = check_add(all_recorder, skirts_filename)
# 找到需要删除的图片名称
all_delete = []
all_filename = tops_filename + pants_filename + skirts_filename
for item in all_recorder:
if item not in all_filename:
all_delete.append(item)
# 分类需要删除的图片
delete_tops = classify_delete(all_delete, tops_recorder)
delete_pants = classify_delete(all_delete, pants_recorder)
delete_skirts = classify_delete(all_delete, skirts_recorder)
return add_tops, add_pants, add_skirts, delete_tops, delete_pants, delete_skirts
def get_color_labels(address_list):
# 创建一个list
color_list = []
# 打开results文件
input_file_0 = open(
'./results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_coat.csv',
'r')
file_to_read_0 = list(csv.reader(input_file_0))
input_file_0.close()
input_file_1 = open(
'./results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_pant.csv',
'r')
file_to_read_1 = list(csv.reader(input_file_1))
input_file_1.close()
input_file_2 = open(
'./results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_skirt.csv',
'r')
file_to_read_2 = list(csv.reader(input_file_2))
input_file_2.close()
for sublist in address_list:
top_color = ''
down_color = ''
for eachline in file_to_read_0:
if sublist[0] == eachline[0]:
top_color = eachline[2]
break
if (sublist[1].split('/')[3]) == 'pant_length_labels':
for eachline in file_to_read_1:
if sublist[1] == eachline[0]:
down_color = eachline[2]
break
if (sublist[1].split('/')[3]) == 'skirt_length_labels':
for eachline in file_to_read_2:
if sublist[1] == eachline[0]:
down_color = eachline[2]
break
sub_color_list = [top_color, down_color]
color_list.append(sub_color_list)
return color_list
def image_composite(imageList):
baseimg = Image.open(imageList[0])
baseimg = baseimg.convert('RGBA')
sz = baseimg.size
basemat = np.atleast_2d(baseimg)
for file in imageList[1:]:
im = Image.open(file, 'r')
im = im.convert('RGBA')
# resize to same width
sz2 = im.size
if sz2 != sz:
im = im.resize((sz[0], round(sz2[0] / sz[0] * sz2[1])),
Image.ANTIALIAS)
mat = np.atleast_2d(im)
basemat = np.append(basemat, mat, axis=0)
report_img = Image.fromarray(basemat)
# report_img.show()
filename = str(uuid.uuid1())
report_img.save('results/' + account.get_signal() + '/matching_results/' +
filename + '.png')
PNG_JPG('results/' + account.get_signal() + '/matching_results/' +
filename + '.png')
return filename + '.jpg'
def get_required_skirt(arrvalue):
try:
required_skirt_list = []
input_file_all_skirt = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_skirt.csv',
'r')
file_to_read_all_skirt = csv.reader(input_file_all_skirt)
all_skirt_rows = [row for row in file_to_read_all_skirt]
for row in all_skirt_rows:
if (float(row[3]) == arrvalue):
required_skirt_list.append(row[0])
return required_skirt_list
except:
print('File Warning')
return []
def prediction_length_classification():
# prediction length results
input_file_0 = open('results/test_length.csv', 'r')
file_to_read_0 = csv.reader(input_file_0)
# prediction_length_coat_temp.csv
output_file_coat = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_coat_temp.csv',
'a',
newline='')
file_to_write_coat = csv.writer(output_file_coat, dialect='excel')
# prediction_length_sleeve_temp.csv
output_file_sleeve = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_sleeve_temp.csv',
'a',
newline='')
file_to_write_sleeve = csv.writer(output_file_sleeve, dialect='excel')
# prediction_length_pant.csv
output_file_pant = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_pant.csv',
'a',
newline='')
file_to_write_pant = csv.writer(output_file_pant, dialect='excel')
# prediction_length_skirt.csv
output_file_skirt = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_skirt.csv',
'a',
newline='')
file_to_write_skirt = csv.writer(output_file_skirt, dialect='excel')
for eachline in file_to_read_0:
result_length = []
result_length.append('datasets/final-rank/' + eachline[0])
result_length.append(eachline[1])
var = eachline[2].split(';')
index = var.index(max(var))
if (eachline[1] == 'coat_length_labels'):
color = color_detect.return_color_2('datasets/final-rank/' +
eachline[0])
result_length.append(color)
if (index <= 1):
result_length.append(1)
elif (index <= 3 and index > 1):
result_length.append(2)
elif (index <= 5 and index > 3):
result_length.append(3)
elif (index <= 7 and index > 5):
result_length.append(4)
file_to_write_coat.writerow(result_length)
elif (eachline[1] == 'sleeve_length_labels'):
result_length.append('')
if (index == 0):
result_length.append(0)
elif (index <= 2 and index > 0):
result_length.append(1)
elif (index <= 4 and index > 2):
result_length.append(2)
elif (index <= 6 and index > 4):
result_length.append(3)
elif (index <= 8 and index > 6):
result_length.append(4)
file_to_write_sleeve.writerow(result_length)
elif (eachline[1] == 'pant_length_labels'):
color = color_detect.return_color_2('datasets/final-rank/' +
eachline[0])
result_length.append(color)
if (index <= 1):
result_length.append(1)
elif (index <= 3 and index > 1):
result_length.append(2)
elif (index <= 5 and index > 3):
result_length.append(3)
file_to_write_pant.writerow(result_length)
elif (eachline[1] == 'skirt_length_labels'):
color = color_detect.return_color_2('datasets/final-rank/' +
eachline[0])
result_length.append(color)
if (index <= 1):
result_length.append(1)
elif (index <= 3 and index > 1):
result_length.append(2)
elif (index <= 5 and index > 3):
result_length.append(3)
file_to_write_skirt.writerow(result_length)
input_file_0.close()
output_file_coat.close()
output_file_sleeve.close()
output_file_pant.close()
output_file_skirt.close()
input_file_coat_temp = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_coat_temp.csv',
'r')
file_to_read_coat_temp = csv.reader(input_file_coat_temp)
input_file_sleeve_temp = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_sleeve_temp.csv',
'r')
file_to_read_sleeve_temp = csv.reader(input_file_sleeve_temp)
output_file_coat = open(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_coat.csv',
'a',
newline='')
file_to_write_coat = csv.writer(output_file_coat, dialect='excel')
rows_coat = [row for row in file_to_read_coat_temp]
rows_sleeve = [row for row in file_to_read_sleeve_temp]
for i in range(len(rows_coat)):
file_to_write_coat.writerow([
rows_coat[i][0], 'coat_length_labels', rows_coat[i][2],
(int(rows_coat[i][3]) + int(rows_sleeve[i][3])) / 2
])
input_file_sleeve_temp.close()
input_file_coat_temp.close()
output_file_coat.close()
os.remove(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_coat_temp.csv'
)
os.remove(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_sleeve_temp.csv'
)
def prehanding():
# Address for downloading images to save
sourDir = 'datasets/final-rank/' + account.get_signal() + '/download/'
destDirCoat = [
'datasets/final-rank/' + account.get_signal() + '/coat_length_labels/',
'datasets/final-rank/' + account.get_signal() +
'/sleeve_length_labels/', 'datasets/final-rank/' +
account.get_signal() + '/collar_design_labels/',
'datasets/final-rank/' + account.get_signal() +
'/lapel_design_labels/',
'datasets/final-rank/' + account.get_signal() + '/neck_design_labels/',
'datasets/final-rank/' + account.get_signal() +
'/neckline_design_labels/'
]
destDirPant = 'datasets/final-rank/' + account.get_signal(
) + '/pant_length_labels/'
destDirSkirt = 'datasets/final-rank/' + account.get_signal(
) + '/skirt_length_labels/'
responseTops = urllib.request.urlopen(urlTops + 'GetTopTXT?phone=' +
account.get_encode_signal())
responsePants = urllib.request.urlopen(urlPants + 'GetPantTXT?phone=' +
account.get_encode_signal())
responseSkirts = urllib.request.urlopen(urlSkirts + 'GetSkirtTXT?phone=' +
account.get_encode_signal())
print(account.get_encode_signal())
print(account.get_signal())
topsfilename = (responseTops.read().decode()).split('\r\n')[:-1]
pantsfilename = (responsePants.read().decode()).split('\r\n')[:-1]
skirtsfilename = (responseSkirts.read().decode()).split('\r\n')[:-1]
print('TopsFilename: {0} {1}'.format(len(topsfilename), topsfilename))
print('PantsFilename: {0} {1}'.format(len(pantsfilename), pantsfilename))
print('SkirtsFilename: {0} {1}'.format(len(skirtsfilename),
skirtsfilename))
add_tops, add_pants, add_skirts, delete_tops, delete_pants, delete_skirts = check_duplication.check(
topsfilename, pantsfilename, skirtsfilename)
print('TopsAdd: {0} {1}'.format(len(add_tops), add_tops))
print('PantsAdd: {0} {1}'.format(len(add_pants), add_pants))
print('SkirtsAdd: {0} {1}'.format(len(add_skirts), add_skirts))
print('TopsDelete: {0} {1}'.format(len(delete_tops), delete_tops))
print('PantsDelete: {0} {1}'.format(len(delete_pants), delete_pants))
print('SkirtsDelete: {0} {1}'.format(len(delete_skirts), delete_skirts))
# delete
delete_all = delete_tops + delete_pants + delete_skirts
if not delete_all:
print('no image needs to delete')
else:
# delete items in recorder
check_duplication.delete_item(
'datasets/final-rank/' + account.get_signal() +
'/download/recorder.csv', delete_all)
# delete items in 3 csv
check_duplication.delete_item(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_coat.csv',
delete_all)
check_duplication.delete_item(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_pant.csv',
delete_all)
check_duplication.delete_item(
'results/' + account.get_signal() +
'/prediction_length_results_classification/prediction_length_skirt.csv',
delete_all)
# delete images in download
for each_to_delete in delete_all:
check_duplication.delete_image('datasets/final-rank/' +
account.get_signal() +
'/download/' + each_to_delete)
for each_dir in destDirCoat + destDirPant + destDirSkirt:
check_duplication.delete_image(each_dir + each_to_delete)
# 写入新增的、需要预测的内容
add_all = add_tops + add_pants + add_skirts
if not add_all:
print('all images has been downloaded')
else:
write_recorder(
add_all, 'datasets/final-rank/' + account.get_signal() +
'/download/recorder.csv', 'a')
for eachDir in destDirCoat:
write_question(eachDir[20:], add_tops, eachDir[35:-1],
'datasets/final-rank/Tests/question.csv', 'a')
write_question(destDirPant[20:], add_pants, destDirPant[35:-1],
'datasets/final-rank/Tests/question.csv', 'a')
write_question(destDirSkirt[20:], add_skirts, destDirSkirt[35:-1],
'datasets/final-rank/Tests/question.csv', 'a')
if add_tops:
for item in add_tops:
item_url = urlTops + 'GetTopPhoto?phone=' + account.get_encode_signal(
) + '&path=' + item
response = urllib.request.urlopen(item_url)
pic = response.read()
with open(sourDir + item, 'wb') as f:
f.write(pic)
for eachDir in destDirCoat:
shutil.copyfile(sourDir + item, eachDir + item)
# write_question_csv(eachDir[20:] + item, eachDir[40:-1])
if add_pants:
for item in add_pants:
item_url = urlTops + 'GetPantPhoto?phone=' + account.get_encode_signal(
) + '&path=' + item
response = urllib.request.urlopen(item_url)
pic = response.read()
with open(sourDir + item, 'wb') as f:
f.write(pic)
shutil.copyfile(sourDir + item, destDirPant + item)
# write_question_csv(destDirPant[20:] + item, destDirPant[40:-1])
if add_skirts:
for item in add_skirts:
item_url = urlTops + 'GetSkirtPhoto?phone=' + account.get_encode_signal(
) + '&path=' + item
response = urllib.request.urlopen(item_url)
pic = response.read()
with open(sourDir + item, 'wb') as f:
f.write(pic)
shutil.copyfile(sourDir + item, destDirSkirt + item)
def start_matching(username, times):
try:
print('start matching...')
match_results_filename = []
weather_list = getWeatherAttribures(weather_url)
if not (weather_list[2]):
weather_list[2] = 0
weather_value = weather_analysis(float(weather_list[1]),
float(weather_list[2]))
print(
'Today Weather: weather = {0[0]}, temperature = {0[1]}, windScale = {0[2]}, humidity = {0[3]}, Weather Value = {1}'
.format(weather_list, weather_value))
basic_top_value = int(weather_value / 2)
basic_bottom_value = weather_value - basic_top_value
matching_up = []
matching_bottom = []
for i in [
basic_top_value - 0.5, basic_top_value, basic_top_value + 0.5,
basic_top_value + 1
]:
if (get_required_coat(i) != []):
matching_up = get_required_coat(i)
for j in [
basic_bottom_value - 1, basic_bottom_value - 0.5,
basic_bottom_value, basic_bottom_value + 0.5
]:
if (get_required_pant(j) != []):
matching_bottom = get_required_pant(j)
if (get_required_skirt(j) != []):
matching_bottom = matching_bottom + get_required_skirt(j)
print('Top Candidate: ', matching_up)
print('Bottom Candidate: ', matching_bottom)
matching_results_color_detect = []
for times in range(times):
temp = matching(matching_up, matching_bottom)
#print('matching result {0}: {1}'.format(times, temp))
matching_results_color_detect.append(temp)
beforeInColor = [
list(t) for t in set(
tuple(_) for _ in matching_results_color_detect)
]
print('before: {0} {1}'.format(len(beforeInColor), beforeInColor))
print('start detecting color matching...')
final_results = color_filter.color_matching(beforeInColor, 4)
for k in range(len(final_results)):
print('composite image {0}: {1}'.format(k, final_results[k]))
match_results_filename.append(image_composite(final_results[k]))
print('matching completed')
print(
'======================================================================================================='
)
print('start uploading...')
image_upload.upload_filename(match_results_filename, username)
for eachpic in match_results_filename:
image_upload.upload_image(
'results/' + account.get_signal() + '/matching_results/' +
eachpic,
eachpic.split('.')[0], username)
print('uploading completed')
print(
'======================================================================================================='
)
except Exception as e:
print(
'======================================================================================================='
)
print('error:', e)
image_upload.upload_filename(['error_image.jpg'], username)
image_upload.upload_image('results/error_image.jpg', 'error_image',
username)