def compare_rgb(path, rgb_folder):
rgb_pic = calculate_pic_rgb(path)
# 主要就是怎么比这几个值了，要根据哪个值为准呢
# 当前我决定取，rgb每个值都差了15 以上，则认为还没加载完成，这个值取啥一样是个难点
r_threshold = 20
g_threshold = 20
b_threshold = 20
conf_default = Config("default.ini")
app_key = conf_default.getconf("default").app
# 陌陌由于白色部分过于多，因此把上限范围弄小点
if app_key == "momo":
r_threshold = 10
g_threshold = 10
b_threshold = 10
# 除了上限，还要加个下限，预防启动黑了的情况
if rgb_pic[0] > rgb_folder[0] + r_threshold:
if rgb_pic[1] > rgb_folder[1] + g_threshold:
if rgb_pic[2] > rgb_folder[2] + b_threshold:
return False
if rgb_pic[0] < rgb_folder[0] - 50:
if rgb_pic[1] < rgb_folder[1] - 50:
if rgb_pic[2] < rgb_folder[2] - 50:
return False
return True
def get_ent_pos(path, out_path, machineName):
origin_height = clip_specific_pic(path,
out_path + machineName + "_clip.jpg")
img = cv2.imread(out_path + machineName + "_clip.jpg")
height, width, something = img.shape
print "height = {}, width = {}".format(height, width)
start_time = datetime.datetime.now()
y_array = [0]
array_flag = False
for i in range(0, height, 2):
# 行数遍历
count = 0
for j in range(0, width, 2):
if img[i, j][0] >= 252 and img[i,
j][1] >= 252 and img[i,
j][2] >= 252:
count += 1
if count >= width / 4:
# print count
if not array_flag:
y_array.append(i)
array_flag = True
else:
if array_flag:
y_array.append(i - 1)
array_flag = False
print "pre-deal time = {}".format(datetime.datetime.now() - start_time)
print y_array
ent_part_pair = []
for t in range(1, len(y_array)):
if y_array[t] - y_array[t - 1] > 120:
# 测试过程可以打开这个
ent_part_pair.append((y_array[t - 1], y_array[t]))
print ent_part_pair
print "all time = {}".format(datetime.datetime.now() - start_time)
x = width * 3 / 4
leng_arr = len(ent_part_pair)
# 特征直播间存起来，似乎不用裁剪，只要专门往上算10个px（有坐标就行），看这10个px平均RGB多少，拿什么存坐标呢（settings）
# 还是要裁剪图片的，因为需要图片来判断是否进入直播间成功
# 裁第几格子，这个要靠app判断，读配置文件
conf_default = Config("default.ini")
app_key = conf_default.getconf("default").app
count = 2
if app_key == "bigo":
count = 1
clip_specific_area(out_path + machineName + "_clip.jpg",
out_path + machineName + "_feature.jpg", width / 4,
ent_part_pair[leng_arr - count][0], width,
ent_part_pair[leng_arr - count][1])
settings.set_value("feature_path", out_path + machineName + "_feature.jpg")
settings.set_value("ent_top_pos_x", x)
settings.set_value(
"ent_top_pos_y",
ent_part_pair[leng_arr - count][0] + origin_height * 0.12)
# 潜规则，要加上12%的高度，因为裁图裁掉了12%
y = (ent_part_pair[leng_arr - count][0] +
ent_part_pair[leng_arr - count][1]) / 2 + origin_height * 0.12
print "finally x = {}, y = {}".format(x, y)
return x, y
def get_device_params():
conf = Config("default.ini")
event = conf.getconf("serial").serial_number
serial = event.split(',')
deviceInfo = DeviceInfo(serial[0])
device_name = deviceInfo.getDeviceInfo(
) # device name 看看韦总到时候怎么处理，把这一个干掉，我这边的操作其实很多余
device_name = re.sub('\s', '', device_name)
return device_name
def startAppBySwipe(self, d, times, name):
conf = Config("default.ini")
app_name = conf.getconf("default").app_name
try:
MLog.info("startAppBySwipe:" + u"try start app ,name = " +
app_name)
bounds = d(text=app_name).info['bounds']
print bounds
except Exception, e:
MLog.info(repr(e))
app_name = "@" + app_name
MLog.info(
u"start_app startAppBySwipe: change app's start name , appname is "
+ app_name)
def get_start_params():
frame = 50
firstLaunchTimes = 0
notFirstLaunchTimes = 0
enterLiveTimes = 1
apkName = u"yy.apk"
package = u"com.duowan.mobile"
try:
MLog.info(u"sys_config get_start_params: 读取配置文件参数...")
conf = Config("default.ini")
frame = conf.getconf("default").frame
firstLaunchTimes = conf.getconf("default").first_start
notFirstLaunchTimes = conf.getconf("default").normal_start
enterLiveTimes = conf.getconf("default").enter_liveroom
apkName = conf.getconf("default").apk_name
package = conf.getconf("default").package
except Exception:
MLog.error(u"获取参数错误,使用默认值")
frame = 50
firstLaunchTimes = 1
notFirstLaunchTimes = 1
enterLiveTimes = 1
apkName = u"yy.apk"
package = u"com.duowan.mobile"
finally:
# start_python 需要运行在init_ffmpeg后面，否则拿不到帧数的值
MLog.info("apkName = " + str(apkName) + " ,first_start = " \
+ str(firstLaunchTimes) + " ,normal_start = " + str(notFirstLaunchTimes) + " ,frame = " + str(frame))
return int(firstLaunchTimes), int(notFirstLaunchTimes), int(
enterLiveTimes), str(apkName), str(package)
def start_calculate(device_name):
conf_default = Config("default.ini")
app_key = conf_default.getconf("default").app
first_launch_result = []
normal_launch_result = []
if app_key == "huya" or app_key == "momo":
first_launch_result = multi_huya_calculate(device_name)
else:
first_launch_result = multi_normal_calculate(device_name, "first")
# 以后想适配虎牙陌陌的话，必须uiautomator那边要手动处理下登录/跳过
normal_launch_result = multi_normal_calculate(device_name, "notfirst")
# TODO 如果进直播间测试次数设置为0，会崩溃，杨帆后续改
try:
enter_ent_result = enter_ent_calculate_new(device_name)
except Exception, e:
MLog.error(u"测试进直播间计算的时候出现崩溃了 + error = ")
MLog.error(traceback.format_exc(e))
enter_ent_result = []
class BaseConfig(object):
def __init__(self):
self.conf = Config("default.ini")
self.param = None
print "BaseConfig init"
def setParams(self, params):
self.param = params
def getParms(self):
return self.param
def getFirstStartTime(self):
if self.param:
return self.param.first_start_times
else:
self.firstTime = self.conf.getconf("default").first_start
return self.firstTime
def getNormalStartTime(self):
if self.param:
return self.param.normal_start_times
else:
self.normalTime = self.conf.getconf("default").normal_start
return self.normalTime
def getEnterLiveRoom(self):
if self.param:
return self.param.enter_liveroom_times
else:
self.enterLiveRomm = self.conf.getconf("default").enter_liveroom
return self.enterLiveRomm
def getApkName(self):
self.apkName = self.conf.getconf("default").apk_name
return self.apkName
def getAppName(self):
if self.param:
return self.param.app_name
else:
self.appName = self.conf.getconf("default").app_name
return self.appName
def getPackage(self):
if self.param:
return self.param.package_name
else:
self.packageName = self.conf.getconf("default").package
return self.packageName
def getFeaturePath(self):
self.featurePath = self.conf.getconf("default").feature_path
return self.featurePath
def getMethod(self):
return self.param.install_method
def multi_huya_calculate_parts(params):
device_name = params["device"] # 设备名称
name_with_suffix = device_name + "_first"
dir_index = params["dir_index"] # 当前算的第几组数据
# rgb_folder = calculate_homepage_rgb() # 计算样本库的rgb均值
conf = Config("apk.ini")
conf_default = Config("default.ini")
app_key = conf_default.getconf("default").app
feature_dir = conf.getconf(app_key).feature # 特征图的文件夹名字
file_count = count_file("./screenrecord/" + name_with_suffix + "/" +
name_with_suffix + "_" + str(dir_index))
real_path = "./screenrecord/" + name_with_suffix + "/" + name_with_suffix + "_" + str(
dir_index) + "/"
real_first_feature_path = path + "/picrepos/feature/" + feature_dir + "/" + device_name + "_launch_feature.jpg"
if not exists(real_first_feature_path):
MLog.debug(
"calculate: first, there is no adapted feature pic for current Phone"
)
real_first_feature_path = path + "/picrepos/feature/" + feature_dir + "/common_launch_feature.jpg"
# 中间launching判断
real_launching_feature_path = path + "/picrepos/feature/" + feature_dir + "/" + device_name + "_launching_feature.jpg"
real_last_feature_path = path + "/picrepos/feature/" + feature_dir + "/" + device_name + "_homepage_feature.jpg"
first = first_frame_find(file_count, real_path,
real_first_feature_path) # 取图片这些步骤好繁琐啊，想想有没办法改进下
launching_index, last = huya_first_find_frame(
file_count, first, real_path, real_launching_feature_path,
real_last_feature_path, rgb_folder) # 虎牙的计算也就只有这句不一样吊
# frame_value = settings.get_value("ffmpeg")
frame_value = 50
total_time = int((last - first + 1) * (1000 / float(frame_value)))
launching_time = int(
(launching_index - first + 1) * (1000 / float(frame_value)))
return dir_index, first, launching_index, last, total_time, launching_time, total_time - launching_time
def sendEmailWithDefaultConfig():
user = u"*****@*****.**"
password = u"lcqctgdcbvklghde"
to_users = u"[email protected], [email protected],[email protected]"
conf = Config("default.ini")
apk_name = conf.getconf("default").apk_name
conf = Config("default.ini")
event = conf.getconf("serial").serial_number
serial = event.split(',')
deviceInfo = DeviceInfo(serial[0])
subject = deviceInfo.getDeviceInfo() + apk_name + u"启动时间数据分析"
content = u"数据分析详见附件："
contentType = u"application/octet-stream"
try:
log_file = make_log_patch()
patchFile = []
MLog.debug(u"sendmail sendEmailWithDefaultConfig: 收集邮件附件:")
for files in os.walk(chart_data_path):
for f in files[2]:
new_file_path = files[0] + f
patchFile.append(new_file_path)
patchFile.append(log_file)
MLog.debug(u"sendmail sendEmailWithDefaultConfig: " + str(patchFile))
except Exception, e:
MLog.error(u"sendmail sendEmailWithDefaultConfig: 收集附件失败! e = " +
repr(e))
patchFile = None
def initConfig(self):
self.isRunning = False
conf = Config("default.ini")
self.sdkPath = conf.getconf("default").sdk_path # sdk路径
self.sdkPathEntry.delete(0, END)
self.sdkPathEntry.insert(0, self.sdkPath)
if self.combox.get() == "自动启动":
self.installMethos = 1
else:
self.installMethos = 2
self.videoPath = ""
self.firstStartTime = conf.getconf("default").first_start # 首次启动次数
self.firstStartEntry.delete(0, END)
self.firstStartEntry.insert(0, self.firstStartTime)
self.normaStartTime = conf.getconf("default").normal_start # 正常启动次数
self.normalStartEntry.delete(0, END)
self.normalStartEntry.insert(0, self.normaStartTime)
self.enterLiveRoonTime = conf.getconf("default").enter_liveroom # 进入直播间次数
self.enterLiveRoomEntry.delete(0, END)
self.enterLiveRoomEntry.insert(0, self.enterLiveRoonTime)
self.appName = conf.getconf("default").app_name # app名称
self.appNameEntry.delete(0, END)
self.appNameEntry.insert(0, self.appName)
self.packageName = conf.getconf("default").package # 包名
self.packageNameEntry.delete(0, END)
self.packageNameEntry.insert(0, self.packageName)
self.appPath = conf.getconf("default").app_path # 安装包地址
self.appPathEntry.delete(0, END)
self.appPathEntry.insert(0, self.appPath)
self.features = "" # 特征图
def calculate_repos_rgb():
conf = Config("apk.ini")
conf_default = Config("default.ini")
app_key = conf_default.getconf("default").app
real_homepage = conf.getconf(app_key).homepage
# homepage_dir = conf.getconf(real_homepage).feature
path = feature_path + "/picrepos/homepage/" + real_homepage + "/"
mean_r = 0
mean_g = 0
mean_b = 0
length_file = base_utils.count_file(feature_path + "/picrepos/homepage/" + real_homepage)
for i in range(1, length_file + 1):
rgb = calculate_pic_rgb(path + base_utils.adapter_num(i) + ".jpg")
mean_r += rgb[0]
mean_g += rgb[1]
mean_b += rgb[2]
mean_r /= length_file
mean_g /= length_file
mean_b /= length_file
MLog.debug("calculate_repos_rgb: the folder mean rgb: r = {}, g = {}, b = {}".format(mean_r, mean_g, mean_b))
return mean_r, mean_g, mean_b
# coding=utf-8
from PIL import Image
import base_utils
# 计算图片的rgb均值
from config.configs import Config
from log.log import MLog
conf = Config("default.ini")
feature_path = conf.getconf("default").feature_path
def calculate_pic_rgb(path):
im = Image.open(path)
pix = im.load()
width = im.size[0]
height = im.size[1]
mean_r = 0
mean_g = 0
mean_b = 0
for x in range(width):
for y in range(height):
r, g, b = pix[x, y]
mean_r += r
mean_g += g
mean_b += b
# print "r = {}, g = {}, b = {}".format(r, g, b)
mean_r /= (width * height)
mean_g /= (width * height)
mean_b /= (width * height)
def getApkName():
conf = Config("default.ini")
apkName = conf.getconf("default").apk_name
return apkName
def main():
# Parse the command line arguments
prog = argparse.ArgumentParser()
prog.add_argument('--model',
type=str,
default='KernelGANFKP',
help='KernelGANFKP or KernelGAN.')
prog.add_argument('--dataset',
'-d',
type=str,
default='../data/datasets/DIV2K/KernelGANFKP_lr_x2',
help='dataset, e.g., DIV2K.')
prog.add_argument(
'--sf',
type=str,
default='2',
help='The wanted SR scale factor, KernelGAN only supports 2 or 4.')
prog.add_argument('--SR',
action='store_true',
default=False,
help='when activated - nonblind SR is performed')
prog.add_argument(
'--real',
action='store_true',
default=False,
help='if the input is real image, to be overwritten automatically.')
prog.add_argument('--path-KP',
type=str,
default='../data/pretrained_models/FKP_x2.pt',
help='path for trained kernel prior')
prog.add_argument('--path-nonblind',
type=str,
default='../data/pretrained_models/usrnet_tiny.pth',
help='path for trained nonblind model')
# to be overwritten automatically
prog.add_argument(
'--input-dir',
'-i',
type=str,
default='../data/datasets/DIV2K/KernelGANFKP_lr_x2',
help='path to image input directory, to be overwritten automatically.')
prog.add_argument(
'--output-dir',
'-o',
type=str,
default='../data/log_KernelGANFKP/DIV2K_KernelGANFKP_lr_x2',
help='path to image output directory, to be overwritten automatically.'
)
prog.add_argument(
'--X4',
action='store_true',
default=False,
help='The wanted SR scale factor, to be overwritten automatically.')
args = prog.parse_args()
# overwritting paths
if args.sf == '2':
args.X4 = False
elif args.sf == '4':
args.X4 = True
else:
print('KernelGAN-FKP only supports X2 and X4')
prog.exit(0)
args.input_dir = '../data/datasets/{}/KernelGANFKP_lr_x{}'.format(
args.dataset, 4 if args.X4 else 2)
args.output_dir = '../data/log_KernelGANFKP/{}_{}_lr_x{}'.format(
args.dataset, args.model, 4 if args.X4 else 2)
# load nonblind model
if args.SR:
netG = USRNet(n_iter=6,
h_nc=32,
in_nc=4,
out_nc=3,
nc=[16, 32, 64, 64],
nb=2,
act_mode="R",
downsample_mode='strideconv',
upsample_mode="convtranspose")
netG.load_state_dict(torch.load(args.path_nonblind), strict=True)
netG.eval()
for key, v in netG.named_parameters():
v.requires_grad = False
netG = netG.cuda()
filesource = os.listdir(os.path.abspath(args.input_dir))
filesource.sort()
for filename in filesource[:]:
print(filename)
# kernel estimation
if args.model == 'KernelGANFKP':
conf = Config_FKP().parse(create_params(filename, args))
kernel = train_FKP(conf)
elif args.model == 'KernelGAN':
conf = Config().parse(create_params(filename, args))
kernel = train(conf)
# nonblind SR
if args.SR:
kernel = map2tensor(kernel)
lr = im2tensor01(read_image(os.path.join(args.input_dir,
filename))).unsqueeze(0)
sr = netG(lr, torch.flip(kernel, [2, 3]), 4 if args.X4 else 2,
(10 if args.real else 0) / 255 *
torch.ones([1, 1, 1, 1]).cuda())
plt.imsave(os.path.join(conf.output_dir_path,
'%s.png' % conf.img_name),
tensor2im01(sr),
vmin=0,
vmax=1.,
dpi=1)
if not conf.verbose and args.SR:
evaluation_dataset(args.input_dir, conf)
prog.exit(0)
def __init__(self):
self.conf = Config("default.ini")
self.param = None
print "BaseConfig init"
def main():
# Parse the command line arguments
prog = argparse.ArgumentParser()
prog.add_argument('--model', type=str, default='DIPFKP', help='models: DIPFKP, DIPSoftmax, DoubleDIP.')
prog.add_argument('--dataset', '-d', type=str, default='Set5',
help='dataset, e.g., Set5.')
prog.add_argument('--sf', type=str, default='2', help='The wanted SR scale factor')
prog.add_argument('--path-nonblind', type=str, default='../data/pretrained_models/usrnet_tiny.pth',
help='path for trained nonblind model')
prog.add_argument('--SR', action='store_true', default=False, help='when activated - nonblind SR is performed')
prog.add_argument('--real', action='store_true', default=False, help='if the input is real image')
# to be overwritten automatically
prog.add_argument('--path-KP', type=str, default='../data/pretrained_models/FKP_x2.pt',
help='path for trained kernel prior')
prog.add_argument('--input-dir', '-i', type=str, default='../data/datasets/Set5/DIPFKP_lr_x2',
help='path to image input directory.')
prog.add_argument('--output-dir', '-o', type=str,
default='../data/log_KernelGANFKP/Set5_DIPFKP_lr_x2', help='path to image output directory')
args = prog.parse_args()
# overwritting paths
args.path_KP = '../data/pretrained_models/FKP_x{}.pt'.format(args.sf)
args.input_dir = '../data/datasets/{}/DIPFKP_lr_x{}'.format(args.dataset, args.sf)
args.output_dir = '../data/log_DIPFKP/{}_{}_lr_x{}'.format(args.dataset, args.model, args.sf)
# load nonblind model
if args.SR:
netG = USRNet(n_iter=6, h_nc=32, in_nc=4, out_nc=3, nc=[16, 32, 64, 64],
nb=2, act_mode="R", downsample_mode='strideconv', upsample_mode="convtranspose")
netG.load_state_dict(torch.load(args.path_nonblind), strict=True)
netG.eval()
for key, v in netG.named_parameters():
v.requires_grad = False
netG = netG.cuda()
filesource = os.listdir(os.path.abspath(args.input_dir))
filesource.sort()
for filename in filesource[:]:
print(filename)
# kernel estimation
conf = Config().parse(create_params(filename, args))
lr_image = im2tensor01(read_image(os.path.join(args.input_dir, filename))).unsqueeze(0)
# crop the image to 960x960 due to memory limit
if 'DIV2K' in args.input_dir:
crop = int(960 / 2 / conf.sf)
lr_image = lr_image[:, :, lr_image.shape[2] // 2 - crop: lr_image.shape[2] // 2 + crop,
lr_image.shape[3] // 2 - crop: lr_image.shape[3] // 2 + crop]
kernel, sr_dip = train(conf, lr_image)
plt.imsave(os.path.join(conf.output_dir_path, '%s.png' % conf.img_name), tensor2im01(sr_dip), vmin=0,
vmax=1., dpi=1)
# nonblind SR
if args.SR:
kernel = map2tensor(kernel)
sr = netG(lr_image, torch.flip(kernel, [2, 3]), int(args.sf),
(10 if args.real else 0) / 255 * torch.ones([1, 1, 1, 1]).cuda())
plt.imsave(os.path.join(conf.output_dir_path, '%s.png' % conf.img_name), tensor2im01(sr), vmin=0,
vmax=1., dpi=1)
if not conf.verbose:
evaluation_dataset(args.input_dir, conf)
prog.exit(0)
import os
try:
from config.configs import Config
except ImportError:
from config.default import DefaultConfig as Config
config = Config()
