SentinelVision is an AI-powered gun detection system designed to enhance security and safety measures, particularly for small businesses vulnerable to armed robberies. The system analyzes video feeds in real-time, detects firearms, and triggers immediate alerts to authorities, enabling swift response and potentially preventing crimes before they escalate.
- Real-Time Gun Detection: Utilizes the YOLOv10 model trained on a dataset of 9,256 annotated images for accurate firearm detection in video streams.
- Enhanced Video Processing: Applies Contrast Limited Adaptive Histogram Equalization (CLAHE) and gamma correction for optimal image quality before analysis.
- Automatic Alert System: Triggers instant alerts to law enforcement upon detecting firearms, along with capturing and forwarding frames for suspect identification.
- User-Friendly Interface: Streamlit-based application for easy video upload and real-time monitoring of detection results.
To install SentinelVision, follow these steps:
-
Clone the repository:
git clone https://github.com/your-username/SentinelVision.git cd SentinelVision -
Install dependencies:
pip install -r requirements.txt
-
Download the YOLOv10 model weights (train9_best.pt) and place them in the root directory of the project.
To run SentinelVision: ```bash streamlit run app.py
