diff --git a/OCR/compare.py b/OCR/compare.py
new file mode 100644
index 0000000..f6a534c
--- /dev/null
+++ b/OCR/compare.py
@@ -0,0 +1,117 @@
+'''
+本方法基于Windows的API
+需要手机投屏/模拟器来操作
+所有的坐标参数都需要按照您自己系统来填写
+本文件只支持比大小
+'''
+
+import os
+
+os.environ["KMP_DUPLICATE_LIB_OK"] = "TRUE"
+
+import cv2
+import numpy as np
+import pyautogui
+from paddleocr import PaddleOCR
+
+import keyboard
+import sys
+import time
+
+# 初始化 PaddleOCR 识别器
+reader = PaddleOCR(use_angle_cls=False, lang='en')
+
+not_found_count = 0
+last_not_found_time = 0
+draw_duration = 0.000
+
+def capture_area():
+    region = (25, 180, 540, 460)  # 识别区域的坐标
+    screenshot = pyautogui.screenshot(region=region)
+    return np.array(screenshot)
+
+def recognize_numbers(image):
+    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
+    # 使用 PaddleOCR 进行文本识别
+    result = reader.ocr(gray, cls=False)
+
+    try:
+        text = ' '.join([item[1][0] for line in result for item in line])  # 提取识别出的文本
+    except:
+        text = ""
+    
+    # 处理识别的文本，提取数字
+    for i in range(len(text.split("\n"))):
+        nums = [int(s) for s in text.split("\n")[i].split() if s.isdigit()]
+        if len(nums) == 3 and nums[1] == 7:
+            return [nums[0], nums[1]]
+        return nums
+
+def draw_comparison(numbers):
+    global not_found_count, last_not_found_time
+
+    if len(numbers) < 2 or len(numbers) and max(numbers) > 100:
+        current_time = time.time()
+        if not_found_count == 0 or current_time - last_not_found_time > 1:
+            not_found_count = 1
+        else:
+            not_found_count += 1
+        last_not_found_time = current_time
+        print("未找到足够的数字进行比较或检测到错误数字")
+
+        if not_found_count >= 10 or len(numbers) and max(numbers) > 100:
+            time.sleep(3)
+            pyautogui.click(268, 534)  # “开心收下”按钮的坐标
+            time.sleep(0.3)
+            pyautogui.click(394, 979)  # “继续”按钮的坐标
+            time.sleep(0.3)
+            pyautogui.click(296, 858)  # “继续PK”按钮的坐标
+            time.sleep(12)
+            
+            print("准备重新开始程序...")
+            time.sleep(0.5)
+            main()
+        return
+
+    first, second = numbers[0], numbers[1]
+    origin_x, origin_y = 180, 650  # 绘制区域的坐标
+    size = 50
+
+
+
+    if first > second:
+        print(f"{first} > {second}")
+        draw_greater_than(origin_x, origin_y, size)
+    elif first < second:
+        print(f"{first} < {second}")
+        draw_less_than(origin_x, origin_y, size)
+
+    not_found_count = 0  
+
+def draw_greater_than(origin_x, origin_y, size):
+    pyautogui.mouseDown(origin_x, origin_y)
+    pyautogui.moveTo(origin_x + size, origin_y + size, duration=draw_duration)
+    pyautogui.moveTo(origin_x, origin_y + size, duration=draw_duration)
+    pyautogui.mouseUp()
+
+def draw_less_than(origin_x, origin_y, size):
+    pyautogui.mouseDown(origin_x + size, origin_y)
+    pyautogui.moveTo(origin_x, origin_y + size, duration=draw_duration)
+    pyautogui.moveTo(origin_x + size, origin_y + size, duration=draw_duration)
+    pyautogui.mouseUp()
+
+def main():
+    keyboard.add_hotkey('-', lambda: sys.exit("进程已结束"))
+
+    try:
+        while True:
+            image = capture_area()
+            numbers = recognize_numbers(image)
+            nums = numbers
+            draw_comparison(nums)
+            time.sleep(0.35)  # 每次绘画及识别的延迟 不要改！
+    except SystemExit as e:
+        print(e)
+
+if __name__ == "__main__":
+    main()
diff --git a/OCR/requirements.txt b/OCR/requirements.txt
new file mode 100644
index 0000000..13db13e
--- /dev/null
+++ b/OCR/requirements.txt
@@ -0,0 +1,4 @@
+opencv-python
+numpy
+pyautogui
+paddleocr
\ No newline at end of file