Wildlife-Protection-Chain is an AI + Blockchain powered project designed to protect endangered animals like tigers, rhinos, and elephants.
Using sample images and videos (Pixabay/Pexels), the system can:
- Detect animals (Tiger, Elephant, Rhino)
- Identify their condition (Normal / Injured)
- Recognize environmental threats (Person, Vehicle, Plastic, Fire, Weapon)
Unlike traditional alert-only systems, Wildlife-Protection-Chain logs every animal detection event on the blockchain including species, condition, environment, time, and location creating a tamper-proof record of wildlife observations.
- Upload Image or Video – User uploads a wildlife photo or clip.
- Processing & Detection – The system analyzes animals, their condition, and surrounding environment.
- DeepSORT Tracking – Animals are tracked across frames for more reliable monitoring.
- Top 5 Key Frames (for videos) – Extracted to let users visually confirm detections and validate AI predictions.
- Voice Alerts – If an injured animal or environmental threat is detected, the system plays a voice alert with the detected threat type and location.
- Blockchain Logging – Each detection event is stored with:
- Timestamp
- Randomized Location (for demo)
- Species (Elephant, Tiger, Rhino)
- Condition (Normal / Injured)
- Threat type(s), if found
- Outputs – Users can download:
- CSV report with all detection events
- Annotated video showing detected animals and threats
- Endangered Wildlife Protection – Elephants, rhinos, and tigers face constant threats from poaching, habitat loss, and human interference.
- Transparency & Trust – Blockchain ensures every record is immutable and cannot be tampered with.
- Human-in-the-loop Validation – Key frame extraction allows people to verify AI outputs instead of relying blindly on automation.
- Sustainability Focused – Scales into conservation tech for long-term environmental monitoring.
- 100% of detections logged on blockchain for transparency.
- Top-5 frame validation ensures trust in AI predictions.
- Automated detection of threats (fire, plastic, vehicles, poachers).
- Real-time voice alerts notify users immediately when an injured animal or threat is found, with threat type and location.
- DeepSORT-based tracking improves consistency across video frames.
- Proof of concept with Pixabay/Pexels datasets.
- Scalable design for drones, ranger cameras, and IoT sensors.
- Note: For best results, upload high-quality images or videos with good lighting. Low-resolution or poorly lit media may affect detection accuracy.
- Frontend / UI: Streamlit
- AI/ML: Computer Vision (YOLO / Custom Image Classification Model + DeepSORT for tracking)
- Blockchain: Custom lightweight blockchain implementation for storing detection events (species, condition, threats, timestamp, location) in a tamper-proof manner
- Media Sources: Pixabay & Pexels (images/videos used for demonstration)
- Drone integration for live monitoring of protected zones using aerial footage.
- Expansion to detect more species beyond elephants, tigers, and rhinos.
- Advanced condition classification: Normal / Injured / Entangled (stuck in nets, ropes, or sharp objects).
- Enhanced environmental threat detection by training on more harmful objects (additional weapons, traps, etc.).
- Behavior-based monitoring: if a species is detected lying down for too long, trigger an immediate rescue alert.
- GPS-based tracking for precise location logging and intervention.
- Decentralized data sharing with NGOs, governments, and researchers.
- Mobile app for rangers to receive real-time blockchain-verified alerts in the field.
Timestamp (s) | Location | Species | Condition | Threats |
---|---|---|---|---|
0.3 | X1-Y9-Z3 | Elephant | Injured | Plastic, Person |
2.5 | A4-B2-C6 | Tiger | Normal | None |
15.7 | D3-E7-F1 | Rhino | Normal | Vehicle, Fire |
- Occasional false detections may occur due to limited demo dataset and variable image quality.
- Tracking (DeepSORT) works but is not perfect in all cases (e.g., occlusions, low light).
- Voice alerts are implemented, but may not be captured in screen recordings due to audio limitations.
- Current implementation is a proof of concept using sample media, not real-time field data.
✨ Wildlife-Protection-Chain demonstrates how AI and Blockchain can work together to create a transparent, scalable, and impactful approach to protecting our planet’s most vulnerable species.