Machine Learning - Recaptcha Solver using YOLO

selenium extract recaptha images

It is not possible to extract the images from reCAPTCHA using Selenium or any other tool because reCAPTCHA is designed to prevent automated extraction of its content.

reCAPTCHA uses various techniques such as image obfuscation and machine learning algorithms to prevent automated programs from extracting its content. Additionally, reCAPTCHA uses a risk analysis engine to determine if a request is coming from a human or a bot, and it can present different types of challenges (such as audio challenges or reCAPTCHA v3) depending on the risk score.

Extracting reCAPTCHA images using automation violates the terms of service of reCAPTCHA and can be considered illegal. It is important to use reCAPTCHA as intended and only for legitimate purposes. If you need to access content that is protected by reCAPTCHA, you should consider using an alternative method or contacting the website owner to request access.

# -*- coding: utf-8 -*-

"""Yolo captch solving.ipynb

Automatically generated by Colaboratory.

Original file is located at

    https://colab.research.google.com/drive/17kG1o_uWwN-KRz9RX_yESdFuqPn9EKLt

"""

!pip install ultralytics

!pip install PIL

!pip install selenium

#input

from ultralytics import YOLO

import cv2

import numpy as np

import matplotlib.pyplot as plt

from matplotlib.image import imread

objective = input("look for ? ")

import os

folder_path = "/content/traffic lights"

jpg_paths = []

for file_name in os.listdir(folder_path):

    if file_name.endswith(".png"):

        jpg_path = os.path.join(folder_path, file_name)

        jpg_paths.append(jpg_path)

print(jpg_paths)

def is_Object_Present(path,objective):

  # Load a model

  model = YOLO("yolov8n.yaml")  # build a new model from scratch

  model = YOLO("yolov8n.pt")  # load a pretrained model (recommended for training)

  # Use the model

  results = model(path)  # predict on an image

  detected_obj_classes = (results[0].boxes.cls).detach().numpy()

  object_classes = results[0][0][0].names 

  result = False

  for objects in detected_obj_classes:

    if object_classes[objects] == objective:

      result = True

  #show image

  img = imread(path)

  plt.imshow(img)

  plt.axis('off')

  plt.show()

  print(result)

  return result

for path in jpg_paths:

  try:

    is_Object_Present(path,objective)

  except:

    pass

# Load a model

model = YOLO("yolov8n.yaml")  # build a new model from scratch

model = YOLO("yolov8n.pt")  # load a pretrained model (recommended for training)

# Use the model

results = model("/content/images/Screenshot 2023-04-27 191430.png")  # predict on an image

detected_obj_classes = (results[0].boxes.cls).detach().numpy()

object_classes = results[0][0][0].names 

unique_object = []

for objects in detected_obj_classes:

  if object_classes[objects] not in unique_object:

    unique_object.append(object_classes[objects])

  # object_set = object_set.union(object_classes[objects])

results[0][0][0].names

img = imread("/content/images/Screenshot 2023-04-27 191430.png")

plt.imshow(img)

plt.axis('off')

plt.show()