【Python游戏】用Python基于centernet在win10平台开发，射击游戏 | 附带源码
相关文件
想学Python的小伙伴可以关注小编的公众号【Python日志】
有很多的资源可以白嫖的哈，不定时会更新一下Python的小知识的哈！！
需要源码的小伙伴可以在公众号回复射击游戏
Python源码、问题解答学习交流群：773162165
开发环境
Python版本：3.7.8
相关模块：
time
mss
numpy
_init_paths
cv2
opts
multiprocessing
以及一些python自带的模块。
环境搭建
安装Python并添加到环境变量，pip安装需要的相关模块即可。
效果展示
代码实现
单线程瞄准射击代码
'''
单线程瞄准射击代码
'''
from __future__ import absolute_import #绝对引用
from __future__ import division
from __future__ import print_function
import torch
import pyautogui
from input_controllers.native_win32_input_controller import NativeWin32InputController
from win32dll_input import Mouse
import mss
import time
import numpy as np
import cv2
import _init_paths
from input_controller import InputController, MouseButton, KeyboardKey, character_keyboard_key_mapping
import os
import cv2
from opts import opts
from detectors.detector_factory import detector_factory
names = ['enemy_0_head', 'enemy_0_body', 'enemy_1_head', 'enemy_1_body', 'enemy_2_head',
'enemy_2_body', 'hostage_0']
test={"force":True}
def draw_res(img,names, results,show_txt=False,enable_bot=False):
heads_loc =[]
body_loc = []
for j in range(1, len(names) + 1):
for bbox in results[j]:
if bbox[4] > 0.3:
bbox = np.array(bbox, dtype=np.int32)
# cat = (int(cat) + 1) % 80
cat = int(j-1)
# print('cat', cat, self.names[cat])
head_c = (0,0,255)
body_c = (255,0,0)
c = (0,255,0)
if not enable_bot:
head_c =c
body_c =c
#print('cat', cat, self.names[cat])
conf = bbox[4]
txt = '{}{:.1f}'.format(names[cat], conf)
font = cv2.FONT_HERSHEY_SIMPLEX
cat_size = cv2.getTextSize(txt, font, 0.5, 2)[0]
if "head" in txt:
img = cv2.rectangle(
img, (bbox[0], bbox[1]), (bbox[2], bbox[3]), head_c, 2)
heads_loc.append((int((bbox[0]+bbox[2])/2),int((bbox[1]+bbox[3])/2)))
elif 'body' in txt:
img = cv2.rectangle(
img, (bbox[0], bbox[1]), (bbox[2], bbox[3]), body_c, 2)
body_loc.append((int((bbox[0]+bbox[2])/2),int((bbox[1]+bbox[3])/2)))
if show_txt:
img = cv2.rectangle(img,
(bbox[0], bbox[1] - cat_size[1] - 2),
(bbox[0] + cat_size[0], bbox[1] - 2), c, -1)
img =cv2.putText(img, txt, (bbox[0], bbox[1] - 2),
font, 0.5, (0, 0, 0), thickness=1, lineType=cv2.LINE_AA)
return img,heads_loc,body_loc
def demo(opt):
os.environ['CUDA_VISIBLE_DEVICES'] = opt.gpus_str
opt.debug = max(opt.debug, 1)
Detector = detector_factory[opt.task]
detector = Detector(opt)
input = NativeWin32InputController()
mouse = Mouse()
detector.pause = False
enable_bot= True
pyautogui.FAILSAFE = False
#mouse.move_mouse((0,0))
pyautogui.move(0,0)
mouse.move_mouse((0,0))
time.sleep(1)
input.click(**test)
input.click(**test)
#pyautogui.click(clicks=2)
with mss.mss() as sct:
# Part of the screen to capture
monitor = {"top": 0, "left": 0, "width": 900, "height": 650}
cv2.namedWindow("OpenCV/Numpy normal")
cv2.moveWindow("OpenCV/Numpy normal", 1020, 10)
pre_pos = (0, 0)
while "Screen capturing":
last_time = time.time() #初始化点位
# Get raw pixels from the screen, save it to a Numpy array
img = np.array(sct.grab(monitor))[:,:,0:3]
ret = detector.run(img)
results = ret['results']
img,heads,bodys = draw_res(img,names,results,show_txt=False,enable_bot=enable_bot)
if enable_bot:
print("bodys",bodys)
for body in bodys:
shot_x = body[0]
shot_y = body[1]
shot_x =int(shot_x/1000.0*800)
shot_y = int(shot_y/750.0*600)
shot = (shot_x,shot_y)
mouse.move_mouse(shot)
mouse.double_click(shot)
#input.click(**test)
#input.click(**test)
#mouse.click(body,button_name='right')
input.click(button=MouseButton.RIGHT,duration=0.05, **test)
#pyautogui.rightClick()
cv2.imshow("OpenCV/Numpy normal", img)
#print('pred time:',ret['tot'])
# Display the picture
print("fps: {}".format(1 / (time.time() - last_time)))
# Press "q" to quit
k =cv2.waitKey(2)
if k == ord("q"):
cv2.destroyAllWindows()
break
elif k== ord("s"):
enable_bot = not enable_bot
if __name__ == '__main__':
args = [
"ctdet",
"--load_model","C:\\MyData\\AI_game\\model_last_res_18.pth",
"--arch","res_18",
"--head_conv","64"
]
opt = opts().init(args=args)
demo(opt)
多线程瞄准-射击代码
'''
多线程瞄准-射击代码
'''
from __future__ import absolute_import #绝对引用
from __future__ import division
from __future__ import print_function
import numpy as np
import mss
import time
import multiprocessing
from multiprocessing import Pipe
from win32dll_input import Mouse
from input_controller import InputController, MouseButton, KeyboardKey, character_keyboard_key_mapping
import os
import cv2
from opts import opts
from detectors.detector_factory import detector_factory
from input_controllers.native_win32_input_controller import NativeWin32InputController
display_time = 2 #每隔dispaly_time显示一次fps
title = "aimbot"
fps =0
pos_scale=1.25 #重要！！！屏幕放大分辨率,游戏鼠标灵敏度必须调为0
names = ['enemy_0_head', 'enemy_0_body', 'enemy_1_head', 'enemy_1_body', 'enemy_2_head',
'enemy_2_body', 'hostage_0']
test={"force":True}
sct = mss.mss()
mouse = Mouse()
calibShotPos = (0,0) #没有目标时在左上角放空枪用于校准射击点
start_time = time.time()
monitor = {"top": 0, "left": 0, "width": int(800*pos_scale)-100, "height": int(600*pos_scale)-50}
args = [
"ctdet",
"--load_model", "C:\\MyData\\AI_game\\model_last_res_18.pth",
"--arch", "res_18",
"--head_conv", "64"
]
opt = opts().init(args=args)
os.environ['CUDA_VISIBLE_DEVICES'] = opt.gpus_str
opt.debug = max(opt.debug, 1)
Detector = detector_factory[opt.task]
detector = Detector(opt)
input = NativeWin32InputController()
mouse = Mouse()
detector.pause = False
def draw_res(img,names, results,show_txt=False,enable_bot=False):
heads_loc =[]
body_loc = []
#cv2.circle(img, shotPos2PixelLoc(calibShotPos), 22, (0, 255, 255), thickness=2) # 用于校正射击点
#cv2.circle(img, shotPos2PixelLoc(calibShotPos2), 22, (0, 255, 255), thickness=2) # 用于校正射击点
for j in range(1, len(names) + 1):
for bbox in results[j]:
if bbox[4] > 0.3:
bbox = np.array(bbox, dtype=np.int32)
# cat = (int(cat) + 1) % 80
cat = int(j-1)
# print('cat', cat, self.names[cat])
head_c = (124,252,0)
body_c = (127,255,212)
c = (0,255,0)
if not enable_bot:
head_c =c
body_c =c
#print('cat', cat, self.names[cat])
conf = bbox[4]
txt = '{}{:.1f}'.format(names[cat], conf)
font = cv2.FONT_HERSHEY_SIMPLEX
cat_size = cv2.getTextSize(txt, font, 0.5, 2)[0]
if "head" in txt:
img = cv2.rectangle(
img, (bbox[0], bbox[1]), (bbox[2], bbox[3]), head_c, 1)
heads_loc.append((int((bbox[0]+bbox[2])/2),int((bbox[1]+bbox[3])/2)))
elif 'body' in txt:
img = cv2.rectangle(
img, (bbox[0], bbox[1]), (bbox[2], bbox[3]), body_c, 1)
body_loc.append((int((bbox[0]+bbox[2])/2),int((bbox[1]+bbox[3])/2)))
if show_txt:
img = cv2.rectangle(img,
(bbox[0], bbox[1] - cat_size[1] - 2),
(bbox[0] + cat_size[0], bbox[1] - 2), c, -1)
img =cv2.putText(img, txt, (bbox[0], bbox[1] - 2),
font, 0.5, (0, 0, 0), thickness=1, lineType=cv2.LINE_AA)
return img,heads_loc,body_loc
def grab_screen(p_input):
while True:
# Get raw pixels from the screen, save it to a Numpy array ,and convert BGRA to BGR
img = np.array(sct.grab(monitor))[:,:,0:3]
p_input.send(img)
def show_screen(p_output2):
global fps, start_time
cv2.namedWindow(title)
cv2.moveWindow(title, 1000, 10)
while True:
img = p_output2.recv()
cv2.imshow(title, img)
fps+=1
TIME = time.time() - start_time
if (TIME) >= display_time :
print("FPS: ", fps / (TIME))
fps = 0
start_time = time.time()
# Press "q" to quit
if cv2.waitKey(3) & 0xFF == ord("q"):
cv2.destroyAllWindows()
break
def shot(target:tuple):
shot_x = target[0]
shot_y = target[1]
shot_x = int(shot_x / pos_scale)
shot_y = int(shot_y / pos_scale)
shot = (shot_x, shot_y)
# print('mouse loc:',shot_x,shot_y)
mouse.move_mouse(shot)
input.click(**test)
input.click(**test)
input.click(button=MouseButton.RIGHT, duration=0.05, **test)
def pixelLoc2shotPos(target:tuple):
shot_x = target[0]
shot_y = target[1]
shot_x = int(shot_x / pos_scale)
shot_y = int(shot_y / pos_scale)
shot = (shot_x, shot_y)
return shot
def shotPos2PixelLoc(pos:tuple):
pos_x = pos[0]
pos_y = pos[1]
loc_x = int(pos_x / pos_scale)
loc_y = int(pos_y /pos_scale)
loc = (loc_x, loc_y)
return loc
def detection(p_output,p_intput2):
mouse.move_mouse((0,0))
input.click(**test)
while True:
img = p_output.recv()
ret = detector.run(img)
results = ret['results']
showimg, heads, bodys = draw_res(img, names, results, show_txt=False, enable_bot=True)
p_intput2.send(showimg)
if len(bodys) is 0:
shot(calibShotPos)
#shot(calibShotPos2)
else:
for body in bodys:
shot(body)
if len(bodys)<2:
for head in heads:
shot(head)
if __name__=="__main__":
p_output,p_input = Pipe()
p_output2,p_input2 = Pipe()
# creating new processes
p1 = multiprocessing.Process(target=grab_screen, args=(p_input, ))
p2 = multiprocessing.Process(target=detection, args=(p_output,p_input2, ))
p3 = multiprocessing.Process(target=show_screen, args=(p_output2, ))
# starting our processes
p1.start()
p2.start()
p3.start()
公众号：Python日志
需要源码的小伙伴可以在公众号回复射击游戏
Python源码、问题解答学习交流群：773162165