Vehicle Recognition and Compilation in Database Software

Abstract

Vehicle Recognition from obtaining images in a motion platform is still challenging. The system would focus and capture attributes of vehicles like color, number plate and speed of the vehicle. The images are being captured from various CCTV systems through distributed intelligence along with time and location stamps. The database used to identify suspects from video clips of crime related CCTV footages. This can be achieved by optical character recognition (OCR) and algorithm based on regression YOLO (You Only Look Once). To recognize an vehicle features, Conda tool is used with Tensor flow and Keras framework.