Cascaded deep learning-based efficient approach for license plate detection and recognition

作者：

Highlights：

• Cascaded deep approach is introduced for license plate detection and recognition.

• Partial labeling is considered for the plate number, city and country regions.

• Three convolutional neural networks models are used independently.

• Arabic characters and northern Iraq region cities recognition are carried out.

• A new challenging license plate dataset for northern Iraq region is introduced.

摘要

•Cascaded deep approach is introduced for license plate detection and recognition.•Partial labeling is considered for the plate number, city and country regions.•Three convolutional neural networks models are used independently.•Arabic characters and northern Iraq region cities recognition are carried out.•A new challenging license plate dataset for northern Iraq region is introduced.

论文关键词：License plate detection and recognition,Deep segmentation,CNN,Arabic number classification

论文评审过程：Received 4 December 2019, Revised 31 December 2019, Accepted 4 February 2020, Available online 13 February 2020, Version of Record 20 February 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.113280