Polyguard stands at the forefront of integrating cutting-edge computer vision technology with Internet of Things (IoT) devices. Our project harnesses the capabilities of YOLOv5 and TensorFlow Lite models, optimized with Coral AI's Accelerator USB for Edge TPU computations on Raspberry Pi 4. Polyguard's mission is to revolutionize surveillance and monitoring systems by introducing high-precision, adaptable person detection technologies.
- Hardware: Raspberry Pi 4 coupled with TensorFlow Lite model and Coral AI USB Accelerator.
- Real-time Detection: Immediate person detection with customizable detection zones via polygon drawing.
- Precision Tracking: Employs bounding box coordinates for pinpoint centroid calculations, ensuring superior detection accuracy.
- Sensitivity Tuning: Fine-tune detection with adjustable centroid ratios, accommodating diverse camera angles.
- User Management: Secure user registration and authentication.
- Camera Configuration: Hassle-free camera setup with unique identifiers.
- Customizable Surveillance: Modify detection zones and sensitivity settings remotely.
- Alert System: SMS notifications upon person detection.
- Footage Review: Convenient video playback for detection events.
- Remote Accessibility: Secure Raspberry Pi exposure through ngrok.
- Robust Security: Integration of advanced security measures for data protection.
- Optimized Performance: Smooth operation at 35-40 FPS with 223x223 frame resolution.
- Raspberry Pi 4: $35
- Raspberry Pi AF Camera 5MP OV5647: $3
- Coral AI USB Accelerator: $59.99
- Total Investment: $97.99
- Programming: Python, Flask.
- Database Management: Firebase.
- Web Development: HTML/CSS with Bootstrap and W3Schools.
Polyguard transcends conventional motion detection systems like MotionEyeOS by leveraging AI-driven vision technology. This approach allows for identifying various objects (e.g., people, animals, vehicles), offering unparalleled flexibility and accuracy in surveillance.
Your expertise can help shape Polyguard's future. We invite contributions and suggest reviewing our guidelines for details on collaborating.
Heartfelt thanks to the dedicated team and community behind Polyguard. Your contributions and skills are the pillars of this innovative project.
