This project demonstrates real-time object recognition using the ResNet50 deep learning model and OpenCV. The ResNet50 model, pre-trained on the ImageNet dataset, is used to predict object classes from video frames captured by a webcam or from a video file.
- Python
- OpenCV
- Keras
- TensorFlow
- Pre-trained ResNet50 model (downloaded automatically by Keras)
-
Clone this repository to your local machine:
git clone https://github.com/yourusername/real-time-object-recognition.git
-
Install the required dependencies:
pip install opencv-python keras tensorflow
-
Run the script:
python object_recognition.py
- When the script is executed, it will access the webcam or the specified video file.
- Each frame from the video source will be processed by the ResNet50 model for object recognition.
- The predicted object class along with its confidence score will be displayed on each frame.
- Press 'q' to exit the application.
Contributions to this project are welcome! Feel free to submit bug reports, feature requests, or pull requests via GitHub.
- Thanks to the developers of OpenCV, Keras, and TensorFlow for their amazing libraries.
- Special thanks to the creators of the ResNet50 model for providing a powerful pre-trained model for object recognition.
For any inquiries or support, please contact jabinjoshua.s@gmail.com
Enjoy real-time object recognition with ResNet50 and OpenCV!