Distributed P2P based Plate Number Classification Architecture for Autonomous Cars in the Cloud Environment

초록

Recently, cloud computing technology has been offering cloud-based plate number classification applications with lower latency. In this paper, we design and implement a new distributed plate number classification system (DPNC). The proposed DPNC system absorbs a more significant number of input sensor data from autonomous cars with a lightweight model that provides high accuracy. In addition, our model has employed the entire convolution network ? Long Shortterm Memory (FCN-LSTM) to predict a total of 3 classes such as image plate, boundary, and number detection. We evaluate the proposed system using an existing Iranian plate dataset containing a collection of plate images using an autonomous car. We used various Amazon cloud services for deploying the proposed DPNC architecture. The experimental results show that the proposed architecture improves end-to-end latency by 2.1 times compared to the traditional architecture.