本来这是去年10月11月做的，当时学算法，看设计模式，搞网站项目太忙了，忘了写，补一下水一篇文章。
去年开始玩赛马娘手游，一款养成类型的爆肝游戏，因为每天都要登录做一下每日任务领奖励，一局游戏操作又重复高，就打算写一个脚本。
在github找到一个项目，但是有些复杂，配置麻烦，还要根据游戏选卡，配装用编写的不同的插件。使用python写的，能在windows上ADB连接手机运行。我想自己或许也能整一个。
查了一个python的图像处理库，在游戏界面的截图，选择几个关键点提取颜色，和给定的预设色计算对比，误差小于一定范围认为在此界面，然后进行操作。
python的代码形式很自由，虽然是脚本，但为了方便编写和阅读，用模块化的函数编程。把主题的养马的对象“马（Uma）”封装起来，用面对对象编程。编写出的代码形式更像是Java或者C++。
import subprocess #子进程管理
from io import BytesIO #从io导入BytesIO模块
import time #
import matplotlib.pyplot as plt #Pyplot 是常用的绘图模块，能很方便让用户绘制 2D 图表。 plt作为matplotlib.pyplot的别名
import numpy as np #科学计算模块，np作为numpy的别名
import random #随机数模块
SAVE_DATA = True #是否保存日志
screenshot = None
# 预先加载要识别的数字
SX0 = plt.imread('icon/0.png')
SX1 = plt.imread('icon/1.png')
SX2 = plt.imread('icon/2.png')
SX3 = plt.imread('icon/3.png')
SX4 = plt.imread('icon/4.png')
SX5 = plt.imread('icon/5.png')
SX6 = plt.imread('icon/6.png')
SX7 = plt.imread('icon/7.png')
SX8 = plt.imread('icon/8.png')
SX9 = plt.imread('icon/9.png')
def get_RB_color(T):
index_0_R = (-T[:, :, 0] + 1) < 0.1
index_0_G = (-T[:, :, 1] + 1) < 0.1
index_0_B = (-T[:, :, 2] + 1) < 0.1
index_0 = index_0_R & index_0_G & index_0_B
index_1_R = abs(-T[:, :, 0] + 0.66) < 0.16
index_1_G = abs(-T[:, :, 1] + 0.36) < 0.18
index_1_B = abs(-T[:, :, 2] + 0.22) < 0.24
index_1 = index_1_R & index_1_G & index_1_B
p = np.zeros(shape=T.shape)
p[index_0, 0] = 1
p[index_1, 1] = 1
return p
B0 = get_RB_color(plt.imread('icon/b0.png'))
B1 = get_RB_color(plt.imread('icon/b1.png'))
B2 = get_RB_color(plt.imread('icon/b2.png'))
B3 = get_RB_color(plt.imread('icon/b3.png'))
B4 = get_RB_color(plt.imread('icon/b4.png'))
B5 = get_RB_color(plt.imread('icon/b5.png'))
B6 = get_RB_color(plt.imread('icon/b6.png'))
B7 = get_RB_color(plt.imread('icon/b7.png'))
B8 = get_RB_color(plt.imread('icon/b8.png'))
B9 = get_RB_color(plt.imread('icon/b9.png'))
class uma():
YaRuKi = 0
TiLi = 0
Friend = False
Health = False
Speed = 0
Stamina = 0
Power = 0
Root = 0
Intellect = 0
Skill_num = 0
def __init__(self):
self.Turns = 0
self.autoLearnSkill = True
self.distance = 0
self.target = [0,0,0,0,0]
self.progress = np.zeros(5)
def get_target(self):
if self.distance == 0:
# 短距离目标
self.target = [999, 400, 999, 400, 400]
elif self.distance == 1:
# 英里目标
self.target = [900, 400, 750, 400, 400]
elif self.distance == 2:
# 中距离目标
self.target = [800, 600, 800, 500, 500]
elif self.distance == 3:
# 长距离目标
#self.target = [800, 800, 750, 350, 300]
self.target = [750, 750, 700, 400, 400]
elif self.distance == 4:
# 自定义目标
self.target = [700, 999, 800, 350, 300]
elif self.distance == 5:
# 自定义目标
self.target = [900, 400, 900, 100, 900]
def cal_target_progress(self):
'''
计算目标属性养成进度
:return:
'''
sx = [self.Speed, self.Stamina, self.Power, self.Root, self.Intellect]
for i in range(5):
if sx[i] < 10:
sx[i] = 1000 #OCR无法识别四位数
self.progress[i] = (sx[i])/self.target[i]
np.set_printoptions(formatter={'float': '{: 0.2f}'.format})
print(f"当前属性进度：\n{self.progress*100}%")
def add_Turns(self):
self.Turns += 1
print("\n===============\nターン：",self.Turns,"\n===============\n")
if SAVE_DATA:
with open('data.txt','a') as f:
data_str = str(self.Turns)+","\
+str(self.TiLi)+","\
+str(self.Speed)+","\
+str(self.Stamina)+","\
+str(self.Power)+","\
+str(self.Root)+","\
+str(self.Intellect)+","\
+str(self.YaRuKi)+"\n"
f.write(data_str)
def get_SX(self):
'''
获取各项属性
:return:
'''
#done
# 速度
SD = cilp_screenshot(100, 1645, 100, 35) #请改为自己手机上对应的区域
self.Speed = SX_number_OCR(SD)
# 耐力
SM = cilp_screenshot(280, 1645, 100, 35)
self.Stamina = SX_number_OCR(SM)
# 力量
LL = cilp_screenshot(450, 1645, 100, 35)
self.Power = SX_number_OCR(LL)
# 根性
GX = cilp_screenshot(620, 1645, 100, 35)
self.Root = SX_number_OCR(GX)
# 智慧
ZH = cilp_screenshot(790, 1645, 100, 35)
self.Intellect = SX_number_OCR(ZH)
def get_TL(self):
'''
获取当前体力
:return:
'''
#done
#p = cilp_screenshot(360, 300, 382, 40) #改为自己手机上的目标区域
p = cilp_screenshot(360, 330, 380, 45)
c1 = p[:, :, 0] == p[:, :, 1]
c2 = p[:, :, 0] == p[:, :, 2]
c3 = p[:, :, 2] == p[:, :, 1]
c = c1 * c2 * c3
# self.TiLi = int((1 - np.sum(c) / (382 * 42)) * 100)
self.TiLi = int((1 - np.sum(c) / (380 * 45)) * 100)
def get_friend(self):
'''
友人事件是否发生
:return:
'''
# c1、c2、c3、c4为4个特征点来识别事件是否反生，对于不同手机，其位置不同，无法共用
# 请自行挑选4个特征点，并更改下面的颜色和位置
c1 = get_color(620, 2056)
t1 = compcolor([255, 63, 121], c1) < 32
c2 = get_color(668, 2046)
t2 = compcolor([255, 69, 131], c2) < 32
c3 = get_color(643, 2028)
t3 = compcolor([255, 255, 255], c3) < 32
c4 = get_color(634, 2071)
t4 = compcolor([255, 255, 255], c4) < 32 if t1 & t2 & t3 & t4: self.Friend = True else: self.Friend = False def get_health(self): ''' 是否生病？ :return: ''' # c1、c2、c3、c4为4个特征点来识别事件是否反生，对于不同手机，其位置不同，无法共用 # 请自行挑选4个特征点，并更改下面的颜色和位置 c1 = get_color(152, 2102) t1 = compcolor([154, 152, 159], c1) > 16
c2 = get_color(365, 2111)
t2 = compcolor([152, 148, 154], c2) > 16
c3 = get_color(181, 2049)
t3 = compcolor([253, 253, 253], c3) < 16
c4 = get_color(303, 2167)
t4 = compcolor([255, 255, 255], c4) < 16 if t1 & t2 & t3 & t4: self.Health = False else: self.Health = True def get_YRK(self): ''' 获取干劲 :return: ''' #done YLK = np.zeros(5) Y1 = [159, 94, 223] # 紫 Y2 = [7, 138, 237] # 蓝 Y3 = [255, 182, 40] # 黄 Y4 = [255, 132, 52] # 橙 Y5 = [249, 73, 130] # 粉红 C = get_color(955, 375) # 请更改坐标 YLK[0] = compcolor(C, Y1) YLK[1] = compcolor(C, Y2) YLK[2] = compcolor(C, Y3) YLK[3] = compcolor(C, Y4) YLK[4] = compcolor(C, Y5) self.YaRuKi = np.argmin(YLK) YRK_text = ["绝不调", "不调", "普通", "好调", "绝好调"] self.YaRuKi_Text = YRK_text[int(self.YaRuKi)] def show_info(self): print("==================") print(" ▲当前体力：",self.TiLi) print(" ▲干劲：", self.YaRuKi_Text) print("------------------") print(" ★速度：",self.Speed) print(" ★耐力：", self.Stamina) print(" ★力量：", self.Power) print(" ★根性：", self.Root) print(" ★智慧：", self.Intellect) print("==================") if self.Friend : print("友人事件已出现！") if not self.Health: print("生病！") def toRest(self): print("当前体力：",self.TiLi,"\n==>休息！")
TAP(167,1900)
PAUSE(1)
TAP(777, 1500)
PAUSE(1.5)
def toGoOut(self):
print("当前干劲：",self.YaRuKi_Text,"\n==>外出！")
TAP(539,2116)
PAUSE(0.3)
TAP(516,902)
PAUSE(0.5)
TAP(777, 1500)
def train(self):
# 自写脚本
# 训练
global index
index = random.randrange(0,4) #随机生成训练
TAP(500, 1840)
PAUSE(0.5)
while True:
if index == 0:
TAP(150, 1900)
PAUSE(0.5)
TAP(150, 1900)
break
if index == 1:
TAP(330, 1900)
PAUSE(0.5)
TAP(330, 1900)
break
if index == 2:
TAP(550, 1900)
PAUSE(0.5)
TAP(550, 1900)
break
if index == 3:
TAP(725, 1900)
PAUSE(0.5)
TAP(725, 1900)
break
if index == 4:
TAP(930, 1900)
PAUSE(0.5)
TAP(930, 1900)
break
def toTore(self):
print("一切正常，==>训练")
deta = random.randint(-20, 20)
TAP(534 + deta, 1904 + deta)
PAUSE(0.5)
CAP()
T = tore()
T.get_all_score()
# 获取目标进度
self.cal_target_progress()
# T.beta = (np.ones(5) - self.progress)*(1-self.Turns/80)*5 + np.ones(5)
T.beta = 50*np.exp(-5*self.progress) + np.ones(5)
if self.Turns <= 11: # 前11轮羁绊多的优先
T.toreninngu(0)
elif self.Turns <= 25: # 12轮-25轮点数多的优先 T.toreninngu(1) elif self.Turns > 25 and self.Turns < 65: # 26轮后改变权重 # T.beta = [0.8,1.4,0.6,0.6,1.0] # T.beta = [1,0.6,1.6,0.3,1.5] T.toreninngu(1) else: # T.beta = [1,1,1,1,1] T.toreninngu(1) def toHospital(self): print("健康异常！=>去保健室")
TAP(259,2018)
PAUSE(1)
TAP(774,1496)
PAUSE(1)
def toLearnSkill(self):
#modify
if not self.autoLearnSkill:
return
print("开始学习前三个技能！")
TAP(900,1840)
PAUSE(1)
TAP(974,1220)
PAUSE(0.2)
TAP(974,1480)
PAUSE(0.2)
TAP(974,1745)
PAUSE(0.2)
TAP(541,1980)#点决定
PAUSE(1)
TAP(769, 2019)#
self.Skill_num += 3
PAUSE(7)
TAP(540,1488)
PAUSE(1)
TAP(130,2210) #点退出
PAUSE(0.3)
#养马算法不用管
class tore():
toSpeed = [0,0,0,0,0]
toStamina = [0,0,0,0,0]
toPower = [0,0,0,0,0]
toRoot = [0,0,0,0,0]
toIntellect = [0,0,0,0,0]
beta = [1,1,1,1,1]
fSpeed = 0
fStamina = 0
fPower = 0
fRoot = 0
fIntellect = 0
def get_up_score(self):
#获取增加属性
score_array = np.zeros(5)
for i in range(5):
T = get_RB_color(cilp_screenshot(30+i*175, 1530, 170, 60))
score_array[i] = B_number_OCR(T)
# print(score_array)
return score_array
def get_all_score(self):
# 速度
CAP()
c = get_color(113, 1868)
T1 = compcolor(c, [218, 218, 218]) < 16*3
c = get_color(175, 1920)
T2 = compcolor(c, [255, 255, 255]) < 16
c = get_color(200, 1895)
T3 = compcolor(c, [160, 186, 245]) <16*3
if(T1&T2&T3):
print("Do Not Touch")
else:
self.tap_points("SD")
CAP()
self.toSpeed=self.get_up_score()
self.fSpeed = self.get_friends_num()
np.set_printoptions(formatter={'float': '{:-5.0f}'.format})
print("？速度：\n属性增加：",sum(self.toSpeed),self.toSpeed,"\n羁绊人数：",self.fSpeed,"人")
# 耐力
self.tap_points("SM")
CAP()
self.toStamina=self.get_up_score()
self.fStamina = self.get_friends_num()
print("？耐力：\n属性增加：",sum(self.toStamina),self.toStamina,"\n羁绊人数：",self.fStamina,"人")
# 力量
self.tap_points("LL")
CAP()
self.toPower=self.get_up_score()
self.fPower = self.get_friends_num()
print("？力量：\n属性增加：",sum(self.toPower),self.toPower,"\n羁绊人数：",self.fPower,"人")
# 根性
self.tap_points("GX")
CAP()
self.toRoot=self.get_up_score()
self.fRoot = self.get_friends_num()
print("？根性：\n属性增加：",sum(self.toRoot),self.toRoot,"\n羁绊人数：",self.fRoot,"人")
# 智慧
self.tap_points("ZH")
CAP()
self.toIntellect=self.get_up_score()
self.fIntellect = self.get_friends_num()
print("？智慧：\n属性增加：",sum(self.toIntellect),self.toIntellect,"\n羁绊人数：",self.fIntellect,"人")
def get_friends_num(self):
#支援卡数目
c0 = [109,108,117]#灰色
c1 = [42,192,255]#淡蓝
c2 = [162,230,29]#浅绿
c3 = [255,172,30]#橙色
c4 = [255,235,120]#淡淡黄
c_base=[81,81,81]
#
for i in range(6,-1,-1):#6-0
c = get_color(901,535+180*i) #
T_base = compcolor(c_base,c) < 32
c = get_color(915,535+180*i)
T0 = compcolor(c0, c) < 16
T1 = compcolor(c1, c) < 16
T2 = compcolor(c2, c) < 16
T3 = compcolor(c3, c) < 16
T4 = compcolor(c4, c) < 16
if T_base & (T0 | T1 | T2 | T3 | T4 ):
return i+1
return 0
def tap_points(self,style):
#正常
deta = random.randint(-20, 20)
if style == "SD":
TAP(156 + deta, 2044 + deta)
elif style == "SM":
TAP(347 + deta, 2025 + deta)
elif style == "LL":
TAP(535 + deta, 2033 + deta)
elif style == "GX":
TAP(730 + deta, 2025 + deta)
elif style == "ZH":
TAP(921 + deta, 2025 + deta)
else:
print("参数错误！！")
PAUSE(0.1)
def toreninngu(self,w):
if w==0:
# 羁绊人数优先
P_array = [self.fSpeed+np.sum(self.toSpeed)/100,
self.fStamina+np.sum(self.toStamina)/100,
self.fPower+np.sum(self.toPower)/100,
self.fRoot+np.sum(self.toRoot)/100,
self.fIntellect+np.sum(self.toIntellect)/100]
print("优先选取羁绊人数更多的训练：\n",P_array)
fmax = max(P_array)
if P_array[0] == fmax:
self.tap_points("SD")
self.tap_points("SD")
elif P_array[1] == fmax:
self.tap_points("SM")
self.tap_points("SM")
elif P_array[2] == fmax:
self.tap_points("LL")
self.tap_points("LL")
elif P_array[3] == fmax:
self.tap_points("GX")
self.tap_points("GX")
elif P_array[4] == fmax:
self.tap_points("ZH")
self.tap_points("ZH")
elif w==1:
# 增加总点数优先（加权）
np.set_printoptions(formatter={'float': '{: 0.3f}'.format})
P_array = [np.sum(self.toSpeed*self.beta),
np.sum(self.toStamina*self.beta),
np.sum(self.toPower*self.beta),
np.sum(self.toRoot*self.beta),
np.sum(self.toIntellect*self.beta)]
print("优先选取加权点数更多的训练")
print("当前加权系数：\n",self.beta)
print("加权结果：\n",P_array)
fmax = max(P_array)
if P_array[0] == fmax:
self.tap_points("SD")
self.tap_points("SD")
elif P_array[1] == fmax:
self.tap_points("SM")
self.tap_points("SM")
elif P_array[2] == fmax:
self.tap_points("LL")
self.tap_points("LL")
elif P_array[3] == fmax:
self.tap_points("GX")
self.tap_points("GX")
elif P_array[4] == fmax:
self.tap_points("ZH")
self.tap_points("ZH")
## 屏幕状态，识别目前在哪个界面，以及点击按钮
class state():
state = 0
#获取状态
def get_state(self):
pass
#分支是否出现
def isChoose(self):
#done
#一个暗灰点，两个选项上的白点，一个亮灰点
c1 = get_color(985, 265)
t1 = compcolor([144, 141, 149], c1) < 32
c2 = get_color(948, 2188)
t2 = compcolor([186, 114, 60], c2) < 32
c3 = get_color(940, 1310)
t3 = compcolor([255, 255, 255], c3) < 32
c4 = get_color(923, 1478)
t4 = compcolor([255, 255, 255], c4) < 32
if t1&t2&t3&t4:
return True
else:
return False
#点击分支选项
def toChoose(self):
#done
print("出现分支选项！")
PAUSE(1)
TAP(940, 1310) #点从下到上数，第二个选项
TAP(940, 1310) #点从下到上数，第二个选项
def isChoose2(self):
#出现两个分支
c1 = get_color(72,1307)
t1 = compcolor([153,219,46],c1) < 32
c2 = get_color(72,1475)
t2 = compcolor([255,205,24],c2) < 32
c3 = get_color(917,1327)
t3 = compcolor([255,255,255],c3) < 32
c4 = get_color(934,1279)
t4 = compcolor([255, 255, 255], c4) < 32
if t1&t2&t3&t4:
return True
else:
return False
def toChoose2(self):
print("出现分支选项！")
TAP(522,1324)
#是否在主菜单，“训练”按键
def isMain(self):
#done
c1 = get_color(400, 1900) #训练蓝色1
t1 = compcolor([44, 127, 212], c1) < 32 #天蓝色
c2 = get_color(670, 1835) #外出白色
t2 = compcolor([244, 154, 136], c2) < 32#纯白
c3 = get_color(650, 1900) #训练蓝色2
t3 = compcolor([71, 145, 219], c3) < 32 #天蓝色
c4 = get_color(850, 250) #目标栏白
t4 = compcolor([255, 255, 255], c4) < 32#纯白
if t1 & t2 & t3 & t4:
return True
else:
return False
def isTargetRace(self):
#育成目标竞赛
#自写函数
#技能按钮蓝条
c1 = get_color(200, 2035)
t1 = compcolor([39, 191, 211],c1) < 32
c2 = get_color(400, 2035)
t2 = compcolor([68, 203, 223],c2) < 32
#竞赛按钮粉红条和2号闸门淡蓝色
c3 = get_color(725, 1975)
t3 = compcolor([182, 225, 255],c3) < 32
c4 = get_color(690, 2020)
t4 = compcolor([255, 89, 144], c4) < 32
if t1&t2&t3&t4:
return True
else:
return False
def goToTargetRace(self):
#去参加育成目标竞赛
#自写函数
print("上场")
TAP(690, 2020) #主界面点击竞赛按钮
PAUSE(1)
TAP(500, 2000) #竞赛界面点击参赛
PAUSE(1)
TAP(800, 1600) #参赛确认弹框，点击参赛按钮
PAUSE(4) #等待加载进围场界面
TAP(400, 2100) #围场界面点击查看结果
PAUSE(10)
TAP(400, 2100) #游戏提示TAP，点击屏幕任意位置
PAUSE(5)
TAP(540, 2100) #竞赛结果点继续
PAUSE(5)
TAP(750, 2100) #弹出奖励点继续
PAUSE(5)
TAP(750, 2150) #目标达成点继续
PAUSE(5)
TAP(550, 2020) #下个目标点继续
def continueChange():
#尝试继续挑战
#自写函数
TAP(930, 1620)
TAP(400, 2100)#回到围场点击查看结果
#粉丝不足
def isGoRace(self):
c1 = get_color(739,1165)
t1 = compcolor([255, 221, 153],c1) < 32
c2 = get_color(123,933)
t2 = compcolor([249,249,249],c2) < 32
c3 = get_color(631,1600)
t3 = compcolor([146, 214, 8],c3) < 32
c4 = get_color(800,1440)
t4 = compcolor([254, 254, 254], c4) < 32
if t1&t2&t3&t4:
return True
else:
return False
#去比赛
def toGoRace(self):
print("目标粉丝不足，参加比赛！")
TAP(767, 1624)
PAUSE(4)
TAP(525,1950)
PAUSE(2)
TAP(800, 1600)
PAUSE(2)
#是否在比赛中
def isRace(self):
#modify
# c1、c2、c3、c4为4个特征点来识别事件是否反生，对于不同手机，其位置不同，无法共用
# 请自行挑选4个特征点，并更改下面的颜色和位置
#技能按钮蓝条
c1 = get_color(200, 2035)
t1 = compcolor([39, 191, 211],c1) < 32
c2 = get_color(400, 2035)
t2 = compcolor([68, 203, 223],c2) < 32
#培育目标白和竞赛按钮粉红条
c3 = get_color(900, 242)
t3 = compcolor([255, 255, 255],c3) < 32
c4 = get_color(690, 2020)
t4 = compcolor([255, 89, 144], c4) < 32
if t1&t2&t3&t4:
return True
else:
return False
#正常去参加比赛
def toRace(self):
#done
print("参加比赛！")
TAP(690, 2020) #主界面点击竞赛按钮
PAUSE(4)
TAP(500, 2000) #竞赛界面点击参赛
PAUSE(1)
TAP(800, 1600) #参赛确认弹框，点击参赛按钮
PAUSE(1)
#是否开始竞赛
def isStartRace(self):
# done
# c1、c2、c3、c4为4个特征点来识别事件是否反生，对于不同手机，其位置不同，无法共用
# 请自行挑选4个特征点，并更改下面的颜色和位置
c1 = get_color(280, 2090)
t1 = compcolor([253, 253, 253], c1) < 32
c2 = get_color(600, 2090)
t2 = compcolor([158, 221, 8], c2) < 32
c3 = get_color(950, 2100)
t3 = compcolor([187, 115, 61], c3) < 32
c4 = get_color(310, 2143)
t4 = compcolor([235, 232, 239], c4) < 32
if t1 & t2 & t3 & t4:
return True
else:
return False
#开始参赛
def toStartRace(self):
#围场
# modify
TAP(400, 2100) #围场界面点击查看结果
PAUSE(10)
TAP(400, 2100) #游戏提示TAP，点击屏幕任意位置
PAUSE(5)
TAP(540, 2100) #竞赛结果点继续
PAUSE(5)
TAP(750, 2150) #目标达成点继续
#是否继承因子
def isInherit(self):
#done
c1 = get_color(430, 1910)
t1 = compcolor([255, 255, 248], c1) < 32
c2 = get_color(488, 1945)
t2 = compcolor([255, 231, 125], c2) < 48
if t1 & t2:
return True
else:
return False
#继承因子
def toInherit(self):
#done
print("因子继承！！")
TAP(535,2001)
#竞赛是否成功
def isRaceSuccess(self):
#modify
# c1、c2、c3、c4为4个特征点来识别事件是否反生，对于不同手机，其位置不同，无法共用
# 请自行挑选4个特征点，并更改下面的颜色和位置
c1 = get_color(510, 2020)
t1 = compcolor([239, 249, 220], c1) < 32
c2 = get_color(415, 2020)
t2 = compcolor([141, 211, 8], c2) < 32
c3 = get_color(890, 1300)
t3 = compcolor([255, 255, 255], c3) < 32
c4 = get_color(430, 750)
t4 = compcolor([255, 255, 137], c4) < 32
if t1 & t2 & t3 & t4:
return True
else:
return False
#竞赛胜利
def toRaceSuccess(self):
#ok
print("目标达成！")
TAP(537, 2090)
PAUSE(3)
TAP(537, 2090)
def isRaceFail(self):
#判断是否育成比赛失败
#自写函数
c1 = get_color(350,730) #绿色继续挑战按钮
t1 = compcolor([124, 203, 12],c1) < 32
c2 = get_color(648,1616) #白色取消按钮
t2 = compcolor([131, 206, 8],c2) < 32
c3 = get_color(328,1615)
t3 = compcolor([121, 64, 22],c3) < 32
c4 = get_color(414,1614)
t4 = compcolor([237, 236, 241], c4) < 32
if t1&t2&t3&t4:
return True
else:
return False
#竞赛失败结束育成
def toRaceFail(self):
#print("目标未达成，结束养成……")
print("目标未达成，闹钟重试》》》")
TAP(785,1600)
#育成是否结束
def isEnd(self):
# c1、c2、c3、c4为4个特征点来识别事件是否反生，对于不同手机，其位置不同，无法共用
# 请自行挑选4个特征点，并更改下面的颜色和位置
c1 = get_color(710, 615)
t1 = compcolor([255, 126, 134], c1) < 32
c2 = get_color(735, 695)
t2 = compcolor([133, 208, 8], c2) < 32
c3 = get_color(395, 1960)
t3 = compcolor([77, 213, 231], c3) < 32
c4 = get_color(640, 1984)
t4 = compcolor([251, 86, 142], c4) < 32 if t1 & t2 & t3 & t4: return True else: return False ## 暂停多少秒 def PAUSE(s): # print("====暂停"+str(s)+"秒====") time.sleep(s) ##封装的ADB命令的函数 def ADB(cmd): result = subprocess.Popen('adb\\adb.exe -s RF8M92C0ZTM '+cmd,shell=True,stdout=subprocess.PIPE) out, err = result.communicate() return out.replace(b'\r\n', b'\n') ##按压这个位置 def TAP(x,y): ADB('shell input tap '+ str(x) +' '+str(y)) #获取屏幕并使用plt绘制一个游戏的小窗口 def CAP(): global screenshot img = ADB('shell screencap -p') picture_stream = BytesIO(img) # 将图片存至内存，减少磁盘读取压力 # screenshot = plt.imread(picture_stream) try: screenshot = plt.imread(picture_stream) except: print("无法获取屏幕信息！") # plt.ion() # plt.clf() # plt.imshow(screenshot) # plt.axis('off') # 去掉坐标轴 # plt.tight_layout(pad=0, h_pad=0, w_pad=0) # plt.pause(0.01) #截屏幕截图 def cilp_screenshot(x,y,w,h): global screenshot return screenshot[y:y+h,x:x+w,0:3] #识别数字 def SX_number_OCR(p): SX = [SX0,SX1,SX2,SX3,SX4,SX5,SX6,SX7,SX8,SX9] num = np.zeros((10,5)) for n in range(0,10): ans = np.zeros(100 - 25) for i in range(0,100-25): A_gry = ( SX[n][:,:,0]+SX[n][:,:,1]+SX[n][:,:,2] )/3 > 0.95
B_gry = ( p[:,i:i+25,0]+p[:,i:i+25,1]+p[:,i:i+25,2] )/3 >0.95
ans[i] = np.sum(A_gry==B_gry)
#print(n,ans)
ii = 0
while np.max(ans) > 830 and ii <4 : ii += 1 num[n,0]=n num[n,ii]=np.argmax(ans) ans[np.argmax(ans)]=0 num_arry = num[:,1:4].T.reshape(-1) SX_number=0 last_max=0 x=0 while np.max(num_arry) > 0:
SX_number += (np.argmax(num_arry)%10)*10**x
x+=1
last_max = np.max(num_arry)
num_arry[np.argmax(num_arry)]=0
while last_max-np.max(num_arry)<20 and last_max>0:
last_max = np.max(num_arry)
num_arry[np.argmax(num_arry)] = 0
# print(SX_number)
return int(SX_number)
#识别数字
def B_number_OCR(p):
B = [B0,B1,B2,B3,B4,B5,B6,B7,B8,B9]
num = np.zeros((10,5))
for n in range(0,10):
ans = np.zeros(170 - 45)
A_0 = B[n][:, :, 0]
A_1 = B[n][:, :, 1]
for i in range(0,170-45):
B_0 = p[:, i:i + 45, 0]
B_1 = p[:, i:i + 45, 1]
ans[i] = 50*(np.sum(A_0 + B_0 == 2)/np.sum(A_0) + np.sum(A_1 + B_1 == 2)/np.sum(A_1))
# print(n,np.max(ans))
ii = 0
while np.max(ans) > 75 and ii < 4 : ii += 1 num[n,0]=n num[n,ii]=np.argmax(ans) ans[np.argmax(ans)]=0 num_arry = num[:,1:4].T.reshape(-1) SX_number=0 last_max=0 x=0 while np.max(num_arry) > 0:
SX_number += (np.argmax(num_arry)%10)*10**x
x+=1
last_max = np.max(num_arry)
num_arry[np.argmax(num_arry)]=0
while last_max-np.max(num_arry)<30 and last_max>0:
last_max = np.max(num_arry)
num_arry[np.argmax(num_arry)] = 0
# print(SX_number)
return int(SX_number)
#获取颜色
def get_color(x,y):
global screenshot
R = int(screenshot[y, x, 0]*256)
G = int(screenshot[y, x, 1]*256)
B = int(screenshot[y, x, 2]*256)
return [R,G,B]
#？
def compcolor(c1,c2):
R = abs(c1[0] - c2[0])
G = abs(c1[1] - c2[1])
B = abs(c1[2] - c2[2])
return R+G+B
if __name__ == "__main__":
print(str(ADB('version'), encoding="utf-8"))
plt.rcParams['toolbar'] = 'None'
fig = plt.figure(figsize=(2.6,5.63))
fig.canvas.manager.set_window_title('ウマ娘自动脚本')
U = uma()
S = state()
U.Turns = int(input("请输入轮次："))
print("\n===============\n短距离：0\n英里：1\n中距离：2\n长距离：3\n===============")
U.distance = int(input("请输入距离："))
U.get_target()
while True:
PAUSE(0.7) #暂停0.7秒 done
CAP() #绘制监视窗口 done
if S.isChoose(): #分支选项 done
S.toChoose() #点击分支选项 done
continue
elif S.isChoose2():
S.toChoose2()
continue
elif S.isGoRace(): #粉丝不足，去参加比赛
U.add_Turns()
S.toGoRace()
continue
elif S.isRace(): #比赛回#done
U.add_Turns()
S.toRace() #正常去参加比赛done
continue
elif S.isStartRace(): #在围场#done
S.toStartRace() #点击开始参赛
continue
elif S.isRaceSuccess(): #竞赛成功#done
S.toRaceSuccess() #竞赛胜利#done
continue
elif S.isRaceFail(): #判断是否育成比赛失败#done
S.toRaceFail() #竞赛失败，闹钟重试 #unknown
continue
elif S.isMain(): #在主菜单 #done
if U.Turns > 48 and U.Skill_num < 1: # 当轮次大于60时开始学习技能 U.toLearnSkill() # 学技能 #unknown continue U.add_Turns() U.get_TL() # 获取体力 #imperfection U.get_YRK() # 获取干劲 #done U.get_SX() # 获取各项属性 #done U.get_health() # 获取健康状况 #done U.show_info() # 显示信息 #done if U.TiLi > 85: # 体力大于85的话优先训练
U.toTore()
continue
# if not U.Health:
# # 生病则去保健室
# U.toHospital()
# continue
# if U.Friend and U.YaRuKi < 4 and U.TiLi < 60: # # 友人事件出现时，优先外出 # U.toGoOut() # continue # if U.Friend and U.Turns > 60 and U.TiLi < 50:
# # 友人事件出现，育成接近尾声时，优先外出
# U.toGoOut()
# continue
if U.TiLi < 50: #体力小于50，去休息
U.toRest()
continue
if U.YaRuKi < 3: #干劲小于4，外出
U.toGoOut()
continue
U.toTore() # 一切正常则去训练
elif S.isInherit(): #继承 #done
S.toInherit()
continue
elif S.isEnd(): #unkown
print("养成结束！")
U.Turns = 0
continue
else:
PAUSE(0.3)
# TAP(750, 2200)
# PAUSE(0.2)
# TAP(537, 2188)
# PAUSE(0.2)
# TAP(537, 2090)
# PAUSE(0.2)
# TAP(756, 2039)
# PAUSE(0.2)
文章评论