Automate WeChat’s “Poke” Feature with Python, Appium, and OpenCV in 30 Lines

Introduction

WeChat’s “punch” (拍一拍) feature can be automated using Python, Appium, and OpenCV. The following tutorial shows how to implement it in about 30 lines of code.

Highlights

Use Appium to control the phone, capture screenshots, locate a friend’s avatar via image recognition, and simulate a double‑tap on the avatar coordinates.

Steps

1. Control the phone with Appium

Install Appium as described elsewhere. The essential capabilities are:

desired_caps = {
    "platformName": "Android",
    "deviceName": "VOG-AL00",
    "appPackage": "com.tencent.mm",
    "appActivity": ".ui.LauncherUI",
    "noReset": "true",
    "fullReset": "false"
}
server = 'http://localhost:4723/wd/hub'
driver = webdriver.Remote(server, desired_caps)

These settings launch WeChat on the connected device.

2. Locate the avatar

We use the Aircv library (a Python wrapper for OpenCV) and its find_template function to find the avatar image within a screenshot.

def matchImg(imgsrc, imgobj, confidencevalue=0.7):
    imsrc = ac.imread(imgsrc)
    imgobj = ac.imread(imgobj)
    match_result = ac.find_template(imsrc, imgobj, confidencevalue)
    return match_result

The function returns a dictionary containing result (center coordinates), rectangle (corner points), and confidence.

3. Capture screenshot and search

driver.save_screenshot('src.png')
imgsrc='src.png'
imgobj='obj.png'
match_result=matchImg(imgsrc, imgobj)

If match_result is None, the avatar was not found and the loop continues.

4. Simulate double‑tap

if match_result is not None:
    result = [match_result['rectangle'][0], match_result['rectangle'][3]]
    for i in range(2):
        driver.tap(result, 100)
    print('拍成功')
else:
    print('没找到头像')

This taps the avatar twice, achieving the “punch” effect.

Conclusion

The script demonstrates a simple way to automate WeChat’s “punch” feature. It can be extended to monitor group chats and automatically “punch” active users or iterate over all members.