Suspicious Object Detection Using AI | ESP32-CAM + Raspberry Pi 5 (Low-Cost, Wearable Idea)

In this project I build a low-cost “smart watch” system that can spot suspicious objects in real time using an ESP32-CAM and a Raspberry Pi 5 with an AI accelerator. The ESP32-CAM acts like a tiny Wi-Fi camera and streams live video, while the Raspberry Pi runs an AI model locally and highlights only the objects we care about (for example, a knife or scissors). The key idea is practical: instead of recording hours of footage and reviewing later, the system can react immediately when something important appears. Because the camera module is small and cheap, you can mount it almost anywhere, and the same concept can even be adapted toward wearable setups where the camera is on a cap or chest mount and the processing unit is carried in a pocket or bag. In the video I show the complete hardware, the live stream feeding into the Pi, and the software that runs detection and displays alerts. If you want the next version, I can add phone notifications, logging, auto clip recording only on events, or a portable battery-powered build.

#ESP32CAM #RaspberryPi5 #EdgeAI #ObjectDetection #IoTSecurity #DIYTech #ComputerVision #Hailo8 #SmartSecurity #MakerProject #CyberPhysicalSystems #OpenSource

Chapters:
0:00 Why this matters (cheap, practical AI safety)
0:10 Hardware overview (ESP32-CAM + Pi 5 + AI add-on)
0:20 Live video stream setup (ESP32 streaming to the Pi)
0:30 Software walkthrough (Python viewer + AI detection)
0:35 Suspicious-only alerts (ignore normal objects)
0:40 Demo test (show detection triggering)
0:45 Wearable/portable idea (small camera, carry the brain)

Видео Suspicious Object Detection Using AI | ESP32-CAM + Raspberry Pi 5 (Low-Cost, Wearable Idea) канала Rai Bahadur Singh