⑩

4_autoTask.py

@@ -0,0 +1,346 @@
+import os,glob
+import time
+import cv2
+from paddleocr import PaddleOCR
+
+import logging
+logging.disable(logging.DEBUG)
+logging.disable(logging.WARNING)
+# 执行 adb 命令函数
+def adb_cmd(cmd):
+    result = os.system(cmd)
+    return result
+# adb 地址
+adb_address = "adb -s emulator-5554"
+# adb 截图
+src_img = adb_address + " exec-out screencap -p > ./resource/image/%DATE:~0,4%%DATE:~8,2%%TIME:~0,2%%TIME:~3,2%%TIME:~6,2%.png" 
+# 获取 当前时间戳 函数
+def to_git_time_str():
+    timestamp = time.time()
+    #将时间戳转换成元组
+    val = time.localtime(timestamp)
+    #将元组转换成字符串
+    time_str= time.strftime("%Y%m%d%H%M%S",val)
+    return time_str
+# 截图 函数
+def to_screen_img():
+    # 获取时间戳
+    res = to_git_time_str()
+    scr_img = adb_address+ " exec-out screencap -p > ./resource/image/{res}.png".format(res=res)
+    adb_cmd(scr_img)
+    # 将 截图 处理成 1280*720
+    path_screen = "./resource/image/"+res+".png"
+    adb_img = cv2.imread(path_screen)
+    adb_img_720p = cv2.resize(adb_img,(1280,720))#将截图720p化
+    return adb_img_720p
+# 图片模板匹配 函数
+def to_macthing_img(temp_img_path):
+    # 目标图片 
+    time.sleep(1)
+    target_img_bgr =  to_screen_img()
+    target_img_gray = cv2.cvtColor(target_img_bgr,cv2.COLOR_BGR2GRAY)
+    target_img = target_img_gray
+    # 模板图片
+    temp_img_bgr = cv2.imread(temp_img_path)
+    target_img_gray = cv2.cvtColor(temp_img_bgr,cv2.COLOR_BGR2GRAY)#灰度处理
+    temp_img = target_img_gray
+
+    # 做一个 颜色匹配 不会写 写修改一下等待时间   
+    # 进行匹配
+    retval = cv2.matchTemplate(target_img,temp_img,method=cv2.TM_CCOEFF_NORMED)
+    minval,maxval,minloc,maxloc = cv2.minMaxLoc(retval)
+    h,w=temp_img.shape[::]
+    return minval,maxval,minloc,maxloc,w,h
+# 点击 函数 - 正常坐标点击
+def click_axis(a,b):
+    x = a*1.5
+    y = b*1.5  
+    click_cmd = adb_address +" shell input tap {x} {y}".format(x=x,y=y)
+    adb_cmd(click_cmd)
+# 点击 函数 - 需要传参
+def click_axis_need(val,w,h):
+    x = int((val[0]+w/2)*1.5)
+    y = int((val[1]+h/2)*1.5)
+    x1,y1=x,y
+    click_cmd = adb_address +" shell input tap {x1} {y1}".format(x1=x1,y1=y1)
+    adb_cmd(click_cmd)    
+# 滑动 函数 - 需要传参数
+def swipe_axis_need(val,w,h,l):
+    # val,w,h 用于计算中心坐标 l 滑动的距离
+    x = int((val[0]+w/2)*1.5)
+    y = int((val[1]+h/2)*1.5)
+    beginx = x 
+    endx = x + l*1.5
+    swipe_cmd = adb_address +" shell input swipe {beginx} {y} {endx} {y}".format(beginx=beginx,y=y,endx=endx)
+    adb_cmd(swipe_cmd)
+
+# 掉落物品匹配函数 
+def to_macthing_drop(target,temp_img_path):
+    # 目标图片 
+    # time.sleep(1)
+    # target_img_bgr =  to_screen_img()
+    target_img_gray = cv2.cvtColor(target,cv2.COLOR_BGR2GRAY)
+    target_img = target_img_gray
+    # 模板图片
+    temp_img_bgr = cv2.imread(temp_img_path)
+    target_img_gray = cv2.cvtColor(temp_img_bgr,cv2.COLOR_BGR2GRAY)#灰度处理
+    temp_img = target_img_gray
+
+    # 做一个 颜色匹配 不会写 写修改一下等待时间   
+    # 进行匹配
+    retval = cv2.matchTemplate(target_img,temp_img,method=cv2.TM_CCOEFF_NORMED)
+    minval,maxval,minloc,maxloc = cv2.minMaxLoc(retval)
+    h,w=temp_img.shape[::]
+    return minval,maxval,minloc,maxloc,w,h
+
+
+
+# 匹配失败 函数
+def match_failed():
+    click_axis(267,39)
+    time.sleep(1)
+    click_axis(94,278)
+    return print("返回首页")
+# 返回值封装 -点击 函数
+def return_click(temp_img):    
+    res1,res2,res3,res4,res5,res6 = to_macthing_img(temp_img)
+    # click_axis_need(res4,res5,res6) if res2 >= 0.9 else match_failed()
+    if res2 >= 0.9:
+        click_axis_need(res4,res5,res6)
+        match_list.append(res2)
+    else:
+        match_list.append(0)  
+        print("未知,匹配失败模板是：",temp_img)     
+#  返回值封装 - 滑动 函数
+def return_swipe(temp_img,l):    
+    res1,res2,res3,res4,res5,res6 = to_macthing_img(temp_img)
+    # click_axis_need(res4,res5,res6) if res2 >= 0.9 else match_failed()
+    if res2 >= 0.9:
+        swipe_axis_need(res4,res5,res6,l)
+        match_list.append(res2)
+    else:
+        match_list.append(0)
+        print("未知,匹配失败模板是：",temp_img)
+# 初始化结果数组
+match_list = [] # 
+counter = 1 # 作战计数器
+soild_num = 0 
+ec_record_num = 0 
+Diketone_num = 0
+Source_rock_num = 0
+lmb_num = 0
+isoiron_debris_num = 0
+gold_num = 0
+Ester_raw_materials_num = 0
+Sugar_substitutes_num = 0
+Broken_devices_num =0
+soild_list=["固源岩掉落情况："] 
+ec_record_list = ["基础作战记录掉落情况："]
+Diketone_list = ["双酮掉落情况："]
+Source_rock_list = ["源岩掉落情况："]
+lmb_list=["龙门币掉落情况："]
+isoiron_debris_list= ["异铁碎片掉落情况："]
+gold_list = ["赤金掉落情况："]
+Ester_raw_materials_list = ["酯原料掉落情况："]
+Sugar_substitutes_list = ["代糖掉落情况："]
+Broken_devices_list = ["破损装置掉落情况："]
+temp_list_str = ["soild","ec_record","Diketone","Source_rock","lmb","isoiron_debris","gold","Ester_raw_materials","Sugar_substitutes","Broken_devices"]
+temp_list_tatal = [soild_list,ec_record_list,Diketone_list,Source_rock_list,lmb_list,isoiron_debris_list,gold_list,Ester_raw_materials_list,Sugar_substitutes_list,Broken_devices_list]
+# 封装 掉落物品 函数
+def drops_handle():
+    # 固源岩 1 2 soild 
+    # 赤金  1 gold
+    # 基础作战记录 1 2 ec_record
+    # 源岩 1  Source_rock
+    # 酯原料 1  Ester_raw_materials
+    # 代糖 1  Sugar_substitutes
+    # 异铁碎片 1   isoiron_debris
+    # 双酮 1 Diketone
+    # 破损装置 1 Broken_devices
+    # 家具  
+    drops = to_screen_img()          
+    for item in temp_list_str:
+        temp_num = item+"_num"
+        temp_list = item+"_list"
+        drops_list_handle(drops,drops,item,exec(temp_num),eval(temp_list))    # exec 是不能给数组使用的 会返回空值
+
+    # 进行封装 传入数组 
+def drops_list_handle(drops,drops_img,temp_img,temp_num,temp_list):
+    res1,res2,res3,res4,res5,res6 = to_macthing_drop(drops,"./resource/temp/{temp_img}.png".format(temp_img=temp_img))
+    if res2 >= 0.9:       
+       top_left = (res4[0], res4[1])
+       centerx =  int(top_left[0] + res5/2)
+       centery =  int(top_left[1] + res6/2)
+       x1 = centerx - 50
+       x2 = centerx + 50
+       y1 = centery- 50
+       y2 = centery+ 50
+       cut_drops = drops_img[y1:y2,x1:x2]
+       temp1 = cv2.imread("./resource/temp/num1.png")
+       temp2 = cv2.imread("./resource/temp/num2.png")
+       temp3 = cv2.imread("./resource/temp/seven.png")
+       res1 = cv2.matchTemplate(cut_drops,temp1,cv2.TM_CCOEFF_NORMED)
+       res2 = cv2.matchTemplate(cut_drops,temp2,cv2.TM_CCOEFF_NORMED)
+       res3 = cv2.matchTemplate(cut_drops,temp3,cv2.TM_CCOEFF_NORMED)
+       minval1,maxval1,minloc1,maxloc1 = cv2.minMaxLoc(res1)
+       minval2,maxval2,minloc2,maxloc2 = cv2.minMaxLoc(res2)
+       minval3,maxval3,minloc3,maxloc3 = cv2.minMaxLoc(res3)
+       if maxval1>=0.9:temp_num=1
+       if maxval2>=0.9:temp_num=2
+       if maxval3>=0.9:temp_num=72
+    #    print(temp_list,temp_num)
+       temp_list.append(temp_num)
+    #    print(temp_img)
+       drop_name =""
+       if temp_img == "soild":
+            drop_name = "固源岩"
+       elif temp_img =="ec_record":
+            drop_name = "基础作战记录"
+       elif temp_img =="Diketone":
+            drop_name = "双酮"
+       elif temp_img =="Source_rock":
+            drop_name = "源岩"
+       elif temp_img =="lmb":
+            drop_name = "龙门币"
+       elif temp_img =="isoiron_debris":
+            drop_name = "异铁碎片"   
+       elif temp_img =="gold":
+            drop_name = "赤金" 
+       elif temp_img =="Ester_raw_materials":
+            drop_name = "酯原料"   
+       elif temp_img =="Sugar_substitutes":
+            drop_name = "代糖"  
+       elif temp_img =="Broken_devices":
+            drop_name = "破损装置"                      
+       else: 
+           print("未匹配到掉落物")   
+       print("检测到 {drop_name} 掉落,数量是".format(drop_name=drop_name),temp_num)
+# 清理文件夹下的所有文件
+def to_clean():
+    for file in glob.glob('D:\\my_demo\\ak-totasks\\resource\\image\\*'):
+        os.remove(file)
+# 递归处理函数
+def to_digui(first_img,end_img):
+    res1,res2,res3,res4,res5,res6 = to_macthing_img(first_img)
+    if res2 >=0.9:
+       print("正常作战中，等待15.5秒继续截图进行模板匹配")
+       time.sleep(15.5)             
+       res1,res2,res3,res4,res5,res6 = to_macthing_img(end_img)
+       if res2>=0.9:                   
+            time.sleep(5)
+            print("作战结束,掉落物品统计-----------------------------------------------------------------------------")
+            # 匹配 固源岩 
+            # 先截图
+            # 封装
+            drops_handle()    
+            return_click(end_img)
+       else : 
+            to_digui(first_img,end_img)
+    # 没有匹配到作战结束 此时在跳转 
+
+# 等待战斗结束 函数
+def await_end(first_img,end_img):    
+   print("等待画面跳转") 
+   time.sleep(10)
+   to_digui(first_img,end_img)
+# 刷完图之后的体力识别
+def reason_judge(x,y,w,h):
+    time.sleep(3)
+    res1,res2,res3,res4,res5,res6= to_macthing_img("./resource/temp/reason_judge.png")
+    if res2>=0.9:
+        # 截图 理智分析 
+       current_reason = to_screen_img()
+       print("作战结束，截图分析理智")
+       # 将图片 进行裁剪 1121,11,78,52 "recommended roi" : [1120,9,80,52]
+       x1=x
+       y1=y
+       x2=x1+w
+       y2=y1+h
+       new_screen = current_reason[y1:y2,x1:x2]
+       ocr = PaddleOCR(use_angle_cls=False,ir_optim=True,use_gpu=False,lang="ch")
+       result = ocr.ocr(new_screen,cls=False,det=False)       
+       # 置信度
+       confidence = result[0][0][1]
+       if confidence >= 0.9:
+           # 识别结果
+            res = int(result[0][0][0].split("/")[0])
+            return res
+       else :
+           print("理智识别失败")
+           return 0
+# 递归函数 循环刷1-7
+def loop_despot(x,y,w,h):
+    time.sleep(2) # 上来先等待5秒
+    res = reason_judge(x,y,w,h)
+    if res == None:
+        print("识别出错了，估计是识别到了黑屏，将理智设置成30，等待下一次识别重新赋值")
+        click_axis(300,200)
+        res = 30
+        print("当前理智：",res)
+    else :
+         print("当前理智：",res)   
+    if res >=6:
+        global counter #定义全局变量 计数器
+        global soild_list
+        global Diketone_list
+        global ec_record_list
+        print("理智大于等于6继续刷图,当前作战次数：",counter) 
+        counter =counter+1  
+        return_click("./resource/temp/start.png")
+        return_click("./resource/temp/start_button.png")
+        await_end("./resource/temp/operation.png","./resource/temp/end_action.png")
+        loop_despot(x,y,w,h)         
+    else:
+       # 计算总掉落情况 - 
+       # 将所有情况全部添加进一个合计list 然后遍历 extend
+       for item in temp_list_tatal:
+           if len(item)>=2:
+                print(item[0])
+                item.pop(0) # 移除第一项
+                res = sum(item)        #    进行求和
+                item.append(str(res))
+                avg = res / counter
+                item.append(avg)
+                print(item)
+       print("理智低于6,停止刷图，总共刷图次数：",counter) 
+       counter = 1  
+def to_despot(): # 刷1-7 函数 
+    # 有无上次作战
+    # 判断体力刷次数
+    # 掉落物品计算
+    # 从首页导航    
+    print("猿神，起洞！")
+    print("清理上一次任务的截图")
+    to_clean()
+    return_click("./resource/temp/fin.png")     #匹配成功进入终端 这里就需要进行判断
+    return_click("./resource/temp/Theme_song.png") # 匹配成功进入主题曲
+    return_click("./resource/temp/disillusionment.png") #匹配成功进入幻灭
+    return_click("./resource/temp/arousal.png")
+    return_swipe("./resource/temp/breathe.png",350)
+    return_swipe("./resource/temp/Co_birth.png",350)
+    return_click("./resource/temp/Dark_Ages.png")
+    return_click("./resource/temp/1_7.png")
+    return_click("./resource/temp/refresh.png")
+    return_click("./resource/temp/one.png")
+    return_click("./resource/temp/start.png")
+    return_click("./resource/temp/start_button.png")
+    await_end("./resource/temp/operation.png","./resource/temp/end_action.png")
+    #刷图结束后 计算一下平均掉落情况
+
+    # 判断理智
+    loop_despot(1140,15,91,51)
+     # 匹配失败 添加一个匹配机制，上述添加一个返回值，如果有一个为false，直接执行匹配失败函数,并将数组清空
+    for res in match_list:
+        if res == 0:
+            match_list.append([])
+            match_failed()
+            print("存在匹配失败！即将返回首页")
+            break
+    #"recommended roi" : [1116,8,84,51]
+# 存源石锭     Source_stone_ingot
+def to_save_source_stone_ingot():
+    print("猿神，起洞！")
+    print("清理上一次任务的截图")
+    return_click("./resource/temp/Enabling_Parameters.png") #"template" : "Enabling_Parameters.png"
+to_despot()
+