Grid Image CAPTCHA: Koordinatenzuordnung und Zellenauswahl

Rasterbild-CAPTCHAs – wie reCAPTCHA v2-Bildherausforderungen – stellen ein 3×3- oder 4×4-Raster dar und fordern Benutzer auf, Zellen auszuwählen, die einer Anweisung entsprechen („Alle Quadrate mit Ampeln auswählen“). CaptchaAI gibt die Zellindizes zurück. In dieser Anleitung erfahren Sie, wie Sie das Raster erfassen, Zellen Koordinaten zuordnen und auf die richtigen Kacheln klicken.

Rasterlayouts

CAPTCHAs verwenden zwei Standardrastergrößen:

3×3 Grid:          4×4 Grid:
1  2  3            1   2   3   4
4  5  6            5   6   7   8
7  8  9            9  10  11  12
                   13  14  15  16

Die Zellen sind von links nach rechts und von oben nach unten nummeriert – Lesereihenfolge.

Schritt 1: Erfassen Sie das Rasterbild

Python (Selen)

import base64
from selenium import webdriver
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC

driver = webdriver.Chrome()
driver.get("https://example.com/form")

# Wait for reCAPTCHA iframe
WebDriverWait(driver, 10).until(
    EC.presence_of_element_located((By.CSS_SELECTOR, "iframe[src*='recaptcha']"))
)

# Switch to challenge iframe
iframes = driver.find_elements(By.CSS_SELECTOR, "iframe[src*='recaptcha']")
challenge_iframe = iframes[-1]  # Challenge iframe is typically the last one
driver.switch_to.frame(challenge_iframe)

# Get the grid image
grid_img = driver.find_element(By.CSS_SELECTOR, "img.rc-image-tile-33, img.rc-image-tile-44")
img_src = grid_img.get_attribute("src")

# Get instruction text
instruction = driver.find_element(
    By.CSS_SELECTOR, ".rc-imageselect-desc-wrapper"
).text
print(f"Instruction: {instruction}")

# Screenshot the grid as base64
img_b64 = grid_img.screenshot_as_base64

# Determine grid size
classes = grid_img.get_attribute("class")
grid_size = "4x4" if "44" in classes else "3x3"
print(f"Grid size: {grid_size}")

driver.switch_to.default_content()

JavaScript (Puppenspieler)

const puppeteer = require('puppeteer');
const fs = require('fs');

const browser = await puppeteer.launch({ headless: false });
const page = await browser.newPage();
await page.goto('https://example.com/form');

// Find the challenge iframe
const frames = page.frames();
const challengeFrame = frames.find(f => f.url().includes('recaptcha'));

// Get instruction
const instruction = await challengeFrame.$eval(
  '.rc-imageselect-desc-wrapper',
  el => el.textContent.trim()
);

// Screenshot the grid image
const gridImg = await challengeFrame.$('img.rc-image-tile-33, img.rc-image-tile-44');
const imgBuffer = await gridImg.screenshot();
const imgBase64 = imgBuffer.toString('base64');

// Determine grid size
const className = await challengeFrame.$eval(
  'img.rc-image-tile-33, img.rc-image-tile-44',
  el => el.className
);
const gridSize = className.includes('44') ? '4x4' : '3x3';
console.log(`Grid: ${gridSize}, Instruction: ${instruction}`);

Schritt 2: An CaptchaAI senden

import requests
import time
import json

API_KEY = "YOUR_API_KEY"

# Parse the instruction to a simple keyword
# "Select all images with traffic lights" → "traffic lights"
import re
keyword_match = re.search(r'(?:with|of|containing)\s+(.+?)\.?$', instruction, re.I)
keyword = keyword_match.group(1) if keyword_match else instruction

# Submit
with open("/tmp/grid.png", "wb") as f:
    f.write(base64.b64decode(img_b64))

with open("/tmp/grid.png", "rb") as f:
    resp = requests.post("https://ocr.captchaai.com/in.php", 
        files={"file": f},
        data={
            "key": API_KEY,
            "method": "post",
            "grid_size": grid_size,
            "img_type": "recaptcha",
            "instructions": keyword,
            "json": "1",
        }
    ).json()

if resp["status"] != 1:
    raise Exception(f"Submit error: {resp['request']}")

task_id = resp["request"]

# Poll
for _ in range(20):
    time.sleep(5)
    result = requests.get("https://ocr.captchaai.com/res.php", params={
        "key": API_KEY, "action": "get", "id": task_id, "json": "1"
    }).json()

    if result["status"] == 1:
        cells = json.loads(result["request"])
        print(f"Cells to click: {cells}")  # e.g., [1, 3, 6, 9]
        break
    if result["request"] != "CAPCHA_NOT_READY":
        raise Exception(f"Error: {result['request']}")

Schritt 3: Ordnen Sie Zellindizes den Klickkoordinaten zu

Konvertieren Sie 1-basierte Zellindizes in Pixelkoordinaten innerhalb des Rasters:

def cell_to_coordinates(cell_index, grid_size, grid_width, grid_height):
    """Convert a 1-based cell index to (x, y) center coordinates."""
    if grid_size == "3x3":
        cols, rows = 3, 3
    else:
        cols, rows = 4, 4

    cell_w = grid_width / cols
    cell_h = grid_height / rows

    # Convert 1-based index to 0-based row/col
    idx = cell_index - 1
    col = idx % cols
    row = idx // cols

    # Center of the cell
    x = col * cell_w + cell_w / 2
    y = row * cell_h + cell_h / 2

    return int(x), int(y)

# Example: grid is 300×300
for cell in cells:
    x, y = cell_to_coordinates(cell, grid_size, 300, 300)
    print(f"Cell {cell} → ({x}, {y})")

Ausgabe für ein 3×3-Raster (300×300):

Cell 1 → (50, 50)
Cell 3 → (250, 50)
Cell 6 → (250, 150)
Cell 9 → (250, 250)

Schritt 4: Klicken Sie auf die Zellen

Selen

from selenium.webdriver.common.action_chains import ActionChains

driver.switch_to.frame(challenge_iframe)

# Get grid element position and size
grid_el = driver.find_element(By.CSS_SELECTOR, ".rc-imageselect-target")
grid_rect = grid_el.rect
grid_w = grid_rect["width"]
grid_h = grid_rect["height"]

actions = ActionChains(driver)

for cell in cells:
    x, y = cell_to_coordinates(cell, grid_size, grid_w, grid_h)
    # Click relative to grid element's top-left corner
    actions.move_to_element_with_offset(
        grid_el,
        x - grid_w / 2,  # offset from center
        y - grid_h / 2
    ).click()

actions.perform()

# Click verify
verify_btn = driver.find_element(By.ID, "recaptcha-verify-button")
verify_btn.click()

driver.switch_to.default_content()

Puppenspieler

// Click each cell by index
const tableRows = await challengeFrame.$$('table.rc-imageselect-table tr');
for (const cellIdx of cells) {
  const row = Math.floor((cellIdx - 1) / (gridSize === '4x4' ? 4 : 3));
  const col = (cellIdx - 1) % (gridSize === '4x4' ? 4 : 3);
  const cell = (await tableRows[row].$$('td'))[col];
  await cell.click();
  await new Promise(r => setTimeout(r, 200));
}

await challengeFrame.click('#recaptcha-verify-button');

Umgang mit dynamischen Kacheln

Einige reCAPTCHA v2-Raster ersetzen angeklickte Kacheln durch neue Bilder. Behandeln Sie dies mit einer Wiederholungsschleife:

def solve_with_dynamic_tiles(driver, api_key, max_rounds=3):
    for round_num in range(max_rounds):
        driver.switch_to.frame(challenge_iframe)

        # Re-capture grid and instruction
        img_b64 = driver.find_element(
            By.CSS_SELECTOR, "img.rc-image-tile-33"
        ).screenshot_as_base64

        # Submit and get cells (same as above)
        cells = submit_and_poll(api_key, img_b64, "3x3", keyword)

        if not cells:
            break

        # Click cells
        click_cells(driver, cells, "3x3")

        # Click verify
        driver.find_element(By.ID, "recaptcha-verify-button").click()

        driver.switch_to.default_content()
        time.sleep(2)

        # Check if solved (no more challenge iframe)
        try:
            driver.switch_to.frame(challenge_iframe)
            driver.switch_to.default_content()
        except Exception:
            return True  # Solved

    return False

Fehlerbehebung

FAQ

Unterstützt CaptchaAI 4×4-Raster?

Ja. Legen Sie grid_size=4x4 fest und die Antwort verwendet die Indizes 1–16.

Wie genau ist die Rasterbildlösung?

Die Genauigkeit hängt von der Bildqualität ab. Senden Sie das Original-CAPTCHA-Bild ohne Zuschneiden oder Komprimieren.

Lösen Sie Rasterbild-CAPTCHAs mit CaptchaAI

Holen Sie sich Ihren API-Schlüssel unter captchaai.com.

Verwandte Leitfäden

• BLS CAPTCHA-Anweisungscode
• Extrahieren von reCAPTCHA-Parametern
• Rasterbild-CAPTCHA automatisch lösen