由于地平线5到目前为止有892辆车，实在过于繁琐复杂，拍卖行因为车辆名字是简写，想靠拍卖蹲车费时费力，况且有些车拍卖行也并不售卖。并且很多网站包括官方的车辆列表，或者其他数据网站，并没有跟拍卖相关的数据，或者存在大量错误（比如年份，名称，车辆）。现在上几乎所有的蹲车软件都是针对单一车的，需要预设值，并不具备真正解决双手，实现全拍卖的功能。
优势:
实现功能:
[color=]（注意：标注DLC 的需要DLC 才能购买，"CAR MODEL LOCATION"=-1 的车不可在拍卖行购买）
每辆车预设蹲30分钟拍卖，如果失败会切换下一辆车，无限循环，直到买齐所有车，脚本才会停止。
内存占用少，约70MB。
前提条件：
预言为英语。（
[color=]理论上任何语言，但是非英语要替换全部images截图）
屏幕分辨率为 1920*1080 （无放大），游戏内 窗口化，特效最低，分辨率1920*1080。
system_setting.png (18.08 KB, 下载次数: 0)
下载附件
2025-1-13 12:56 上传
video_setting.png (165.47 KB, 下载次数: 0)
下载附件
2025-1-13 12:56 上传
graphics_setting.png (291.25 KB, 下载次数: 0)
下载附件
2025-1-13 12:56 上传
Excel 文件（FH5_all_cars_info_v3.xlsx） 解释：
[color=]一般情况下不需要调整，除非有新车加入打乱顺序
car_make_order.PNG (118.65 KB, 下载次数: 0)
下载附件
2025-1-13 12:56 上传
使用方法及程序界面：
auction_house.png (883.15 KB, 下载次数: 0)
下载附件
2025-1-13 12:30 上传
auction_house.PNG (113.44 KB, 下载次数: 0)
下载附件
2025-1-13 12:56 上传
程序界面：
捕获.PNG (68.02 KB, 下载次数: 0)
下载附件
2025-1-13 13:26 上传
下载：
全部源码（包含OCR识别892全部车获取数据等源码）: Github
成品下载：Github
892全部车获取数据下载：
FH5_all_cars_info_v3.zip
(51.78 KB, 下载次数: 2)
2025-1-13 14:05 上传
点击文件名下载附件
下载积分: 吾爱币 -1 CB
与主体源码匹配的images 文件夹：
images.zip
(76.77 KB, 下载次数: 1)
2025-1-13 14:07 上传
点击文件名下载附件
下载积分: 吾爱币 -1 CB
感谢热心坛友提供的蓝奏云：https://wwiy.lanzoub.com/i85kg2kuwp9e
主体源码（下载892车全部数据和images文件夹，放入同一文件夹下即可运行）：
[Python] 纯文本查看 复制代码import os
import cv2
import time
import numpy as np
import pandas as pd
from openpyxl import load_workbook
import pyautogui as pyau
import pydirectinput as pydi
import pygetwindow as gw
def find_image_with_percentage(image_path, region=None, width_ratio=1, height_ratio=1):
# Take a screenshot and convert it to a format suitable for OpenCV
screenshot = pyau.screenshot(region=region)
screenshot_cv = np.array(screenshot)
screenshot_cv = cv2.cvtColor(screenshot_cv, cv2.COLOR_RGB2BGR)
template = cv2.imread(image_path, cv2.IMREAD_COLOR)
screenshot_cv = cv2.resize(screenshot_cv, (int(screenshot_cv.shape[1]/width_ratio), int(screenshot_cv.shape[0]/height_ratio)))
result = cv2.matchTemplate(screenshot_cv, template, cv2.TM_CCOEFF_NORMED)
return result
def find_max_percentage_image(image_path, region=None, width_ratio=1, height_ratio=1, threshold=0.8):
""" This function find the location with the highest probability matching the given image
"""
if type(image_path) == str:
image_path = [image_path]
return_index = True if len(image_path) > 1 else False
_index = 0
_error = 5*width_ratio
best_prob = 0
best_loc = ()
for each_image_path in image_path:
result = find_image_with_percentage(each_image_path, region=region, width_ratio=width_ratio, height_ratio=height_ratio)
loc = np.where(result >= threshold)
for pt in zip(*loc[::-1]):
added= True
if added and best_prob 0:
multi_press(button1, _times)
else:
multi_press(button2, abs(_times))
def go_to_MAKE(old_make, old_CAR_MODEL_ORDER, new_make, new_CAR_MODEL_ORDER):
x, y = np.array(old_make) - np.array(new_make)
if x!=0 or y!=0:
pydi.press('enter')
time.sleep(0.3)
multi_press_cond('a', 'd', x)
multi_press_cond('w', 's', y)
pydi.press('enter')
time.sleep(0.3)
multi_press('s', 1)
if x ==0 and y == 0: # same brand
CAR_MODEL_move = new_CAR_MODEL_ORDER-old_CAR_MODEL_ORDER
else:
CAR_MODEL_move = new_CAR_MODEL_ORDER
multi_press_cond('d', 'a', CAR_MODEL_move)
multi_press('s', 5)
pydi.press('enter')
def measure_game_window(title):
""" Measure the game window size
"""
try:
game_window = gw.getWindowsWithTitle(title)[0]
if game_window:
if not game_window.isMinimized:
game_window.restore()
game_window.activate()
# measure game window
left, top, width, height = game_window.left, game_window.top, game_window.width, game_window.height
print(f"Game window found: {title}")
print(f"Size '{title}' is {width}x{height} pixels.")
return left, top, width, height
else:
print("Game window not found. Check the title.")
except IndexError:
print("Game window not found. Make sure the title is correct.")
except Exception as e:
print(f"An error occurred: {e}")
def press_image(image_path, search_region, width_ratio, height_ratio, threshold):
best_loc = find_max_percentage_image(image_path, search_region, width_ratio, height_ratio, threshold)
left, top, width, height = search_region
if best_loc:
pydi.press('enter')
return True
return False
def click_left():
# Simulate the left mouse click
pydi.mouseDown()
time.sleep(0.05)
pydi.mouseUp()
def write_excel(data, output_path, sheet_name):
# write dataframe to excel and format the excel
data.to_excel(output_path, index=False, sheet_name=sheet_name)
workbook = load_workbook(output_path)
sheet = workbook.active
sheet.auto_filter.ref = sheet.dimensions
for col in sheet.columns:
max_length = 0
col_letter = col[0].column_letter
for cell in col:
try:
# Calculate the maximum length of values in the column (including the header)
max_length = max(max_length, len(str(cell.value)))
except:
pass
# Adjust the column width
sheet.column_dimensions[col_letter].width = max_length + 2
# Save the workbook with adjusted widths
workbook.save(output_path)
def main():
threshold=0.8
game_title = "Forza Horizon 5"
# screenshot regions here
left, top, width, height = measure_game_window(game_title)
width_ratio, height_ratio = 1, 1
search_region_auction = (230+left, 590+top, 910, 310)
search_region_carpage = (790+left, 190+top, 810, 90)
search_region_bid = (520+left, 380+top, 610, 100)
search_region_carpage2 = (60+left, 165+top, 180, 40)
current_directory = os.getcwd()
# screenshots here
image_path_SA = current_directory + '/images/SA.png'
image_path_CF = current_directory + '/images/CF.png'
image_path_AT = current_directory + '/images/AT.png'
image_path_BF = current_directory + '/images/BF.png'
image_path_PB = current_directory + '/images/PB.png'
image_path_BS = current_directory + '/images/BS.png'
image_path_NB = current_directory + '/images/NB.png'
image_path_VS = current_directory + '/images/VS.png'
image_path_AO = current_directory + '/images/AO.png'
# car info details here
car_info_file_path = "./FH5_all_cars_info_v3.xlsx"
car_sheet_name = 'all_cars_info'
print('Welcome to the Forza 5 CAR BUYOUT Snipper')
print('The script will start in 5 seconds')
time.sleep(5)
print('Starts')
change_make = True
new_make, new_CAR_MODEL_ORDER = (0,0), 0
missed_match_times = 0
start_time, all_snipe_index, failed_snipe = 0, [], False
while True:
end_time = time.time()
if end_time - start_time > 1800:
change_make = True
failed_snipe = True
vertify_press_SA = press_image(image_path_SA, search_region_auction, width_ratio, height_ratio, threshold)
time.sleep(0.1)
# change car here
if change_make and find_max_percentage_image(image_path_CF, search_region_auction, width_ratio, height_ratio, threshold):
if failed_snipe:
failed_snipe=False
print('TIME OUT, Switching to Next Auction Sniper!')
start_time = time.time()
vertify_press_CF = True
change_make = False
# read file and filter non-zero cars
df = pd.read_excel(car_info_file_path, car_sheet_name)
if len(df[df['BUYOUT NUM'] > 0]) == 0:
print('Finish Sniping!')
break
all_snipe_index = df[df['BUYOUT NUM'] > 0].index.tolist() if all_snipe_index == [] else all_snipe_index
index = all_snipe_index.pop()
old_make, old_CAR_MODEL_ORDER = new_make, new_CAR_MODEL_ORDER
CAR_MAKE,CAR_MAKE_LOCATION, CAR_MODEL_Full_Name, CAR_MODEL_Short_Name, CAR_MODEL_LOCATION = df.iloc[index,].values[:5]
new_make, new_CAR_MODEL_ORDER = eval(CAR_MAKE_LOCATION), CAR_MODEL_LOCATION
print('Sniping',CAR_MAKE, CAR_MODEL_Full_Name)
# modify car details
multi_press('w', 6) # one more move make sure it goes smoothly
go_to_MAKE(old_make, old_CAR_MODEL_ORDER, new_make, new_CAR_MODEL_ORDER)
else:
vertify_press_CF = press_image(image_path_CF, search_region_auction, width_ratio, height_ratio, threshold)
time.sleep(0.5)
found_carpage = find_max_percentage_image(image_path_AT, search_region_carpage, width_ratio, height_ratio, threshold)
# if found car in stock
if found_carpage:
stop = False
while not stop:
time.sleep(0.1)
pydi.press('y')
# detect whether we can place bid, if not it means we either missed it or still loading
found_bid = find_max_percentage_image(image_path_PB, search_region_bid, width_ratio, height_ratio, threshold)
found_outbid = find_max_percentage_image(image_path_VS, search_region_bid, width_ratio, height_ratio, threshold)
found_auction_option = find_max_percentage_image(image_path_AO, search_region_bid, width_ratio, height_ratio, threshold)
if found_bid or found_outbid or found_auction_option: stop = True
# if found buy option
if found_bid:
pydi.press('s')
pydi.press('enter')
pydi.press('enter')
time.sleep(5)
stop = False
# get the bid result
while not stop:
found_buyoutfail = find_max_percentage_image(image_path_BF, search_region_bid, width_ratio, height_ratio, threshold)
found_buyoutsuccess = find_max_percentage_image(image_path_BS, search_region_bid, width_ratio, height_ratio, threshold)
if found_buyoutfail:
print('BUYOUT Failed!')
pydi.press('enter')
pydi.press('esc')
stop = True
if found_buyoutsuccess:
print('BUYOUT Success!')
# Change to the next car
df.loc[index, 'BUYOUT NUM'] = df['BUYOUT NUM'][index]-1
write_excel(df, car_info_file_path, car_sheet_name)
if df.loc[index, 'BUYOUT NUM'] == 0:
change_make = True
old_make, old_CAR_MODEL_ORDER = new_make, new_CAR_MODEL_ORDER
pydi.press('enter')
pydi.press('esc')
stop = True
else:
print('BUYOUT Missed!')
pydi.press('esc')
time.sleep(0.1)
# if stuck somewhere unknown, this may work
if vertify_press_SA==False and vertify_press_CF==False and found_carpage==None:
pyau.moveTo(left+20, top+20, duration=0.001)
click_left()
if missed_match_times >=2:
pydi.press('esc')
missed_match_times += 1
else:
missed_match_times = 0
# return to main auction page
if found_carpage or vertify_press_CF or find_max_percentage_image(image_path_NB, search_region_carpage2, width_ratio, height_ratio, threshold):
pydi.press('esc')
time.sleep(0.4)
if __name__ == "__main__":
main()
