Loading repository data…
Loading repository data…
BernardNyarko / repository
With recent advances in both Artificial Intelligence (AI) and Internet of Things (IoT) capabilities, it is more possible than ever to implement surveillance systems that can automatically identify people who might represent a potential security threat to the public in real-time. Imagine a surveillance camera system that can detect various on-body weapons, suspicious objects, and traffic. This system could transform surveillance cameras from passive sentries into active observers, which would help prevent a possible mass shooting in a school, stadium, or mall. In this project, we tried to realize such systems by implementing Smart-Monitor, an AI-powered threat detector for intelligent surveillance cameras. The developed system can be deployed locally on the surveillance cameras at the network edge. Deploying AI-enabled surveillance applications at the edge enables the initial analysis of the captured images on-site, reducing the communication overheads and enabling swift security actions. We developed a mobile app that users can detect suspicious objects in an image and video captured by several cameras at the network edge. Also, the model can generate a high-quality segmentation mask for each object instance in the photo, along with the confidence percentage. The camera side used a Raspberry Pi 4 device, Neural Compute Stick 2 (NCS 2), Logitech C920 webcam, motion sensors, buzzers, pushbuttons, LED lights, Python Face recognition, and TensorFlow Custom Object Detection. When the system detects a motion in the surrounding environment, the motion sensors send a signal to the Raspberry Pi device notifying it to start capturing images for such physical activity. Using Python’s face recognition and TensorFlow 2 custom object detection Smart-Monitor can recognize eight classes, including a baseball bat, bird, cat, dog, gun, hammer, knife, and human faces. Finally, we evaluated our system using various performance metrics such as classification time and accuracy, scalability, etc.
With recent advances in both Artificial Intelligence (AI) and Internet of Things (IoT) capabilities, it is more possible than ever to implement surveillance systems that can automatically identify people who might represent a potential security threat to the public in real-time. Imagine a surveillance camera system that can detect various on-body weapons, suspicious objects, and traffic. This system could transform surveillance cameras from passive sentries into active observers, which would help prevent a possible mass shooting in a school, stadium, or mall. In this project, we tried to realize such systems by implementing Smart-Monitor, an AI-powered threat detector for intelligent surveillance cameras. The developed system can be deployed locally on the surveillance cameras at the network edge. Deploying AI-enabled surveillance applications at the edge enables the initial analysis of the captured images on-site, reducing the communication overheads and enabling swift security actions. We developed a mobile app that users can detect suspicious objects in an image and video captured by several cameras at the network edge. Also, the model can generate a high-quality segmentation mask for each object instance in the photo, along with the confidence percentage. The camera side used a Raspberry Pi 4 device, Neural Compute Stick 2 (NCS 2), Logitech C920 webcam, motion sensors, buzzers, pushbuttons, LED lights, Python Face recognition, and TensorFlow Custom Object Detection. When the system detects a motion in the surrounding environment, the motion sensors send a signal to the Raspberry Pi device notifying it to start capturing images for such physical activity. Using Python’s face recognition and TensorFlow 2 custom object detection Smart-Monitor can recognize eight classes, including a baseball bat, bird, cat, dog, gun, hammer, knife, and human faces. Finally, we evaluated our system using various performance metrics such as classification time and accuracy, scalability, etc.