الفهرس | Only 14 pages are availabe for public view |
Abstract License Plate Recognition (LPR) is an important function in intelligent traffic control systems. The main goal of this work is to make segmentation faster using parallel implementation of algorithms on GPU. We use the GPU to accelerate LPR performance where a large number of cores collaborate in achieving a real time performance of LPR. GPUs are emerging as powerful parallel frameworks at a less expensive cost. Our work employs extensive usage of multi-cored GPU. An efficient use of Compute Unified Device Architecture (CUDA) is to implement our algorithms on GPU. Experimental results show that parallel functioning when doing on a GPU is important and effective. |