MOVING OBJECT DETECTION & ITS REAL TIME IMPLEMENTATION USING DEEP LEARNING

Abstract

AI is a rapidly expanding field. Because of its ease of use, availability, and great performance. Deep learning on low-cost computers like the Raspberry Pi may be employed to achieve results. Detecting the objects in a hall or pathways can give an advantage in many ways such as security, monitoring and much more. In the context it can provide an Artificial replicated image through other senses by observing the environment through camera, to see if the throughput is sufficient for real-time object identification, we need some sort of algorithm that detect and recognize objects and tag them accordingly. The computer Vision provides the path to object detection and AI in this area. YOLO (You Only Look Once) change the whole concept by coming up with the model that uses linear regression as base in order to detect objects and provide much higher speed and accuracy. In this research pre-trained Y0L05 model is used to detect objects on a raspberry pi model 4B board that is a single board computer which provides good processing power with low power consumption. One of the most important application using Computer Vision is Object detection that allows detecting instances ofreal time stream of images or static image. This allows to identify and locate object within the frame. To provide efficient object detection using raspberry pi is the aim of our project. Raspberry pi provides portability and power efficiency which makes this model a handy and environment friendly tool for monitoring and much more