access icon free Evaluation methodology for license plate recognition systems and experimental results

Two of the most critical requirements for the development of license plate recognition (LPR) systems are a comprehensive database of license plate images and a testing procedure to evaluate the system. This study addresses both issues through the LPR system test (LPRST), which establishes the first performance benchmark for the LPR system. There are three important parts in the study: (i) the comprehensive, structured and labelled license plate image database is established, which is openly available to researchers; (ii) the first technical evaluation protocol for the LPR system is established, in which the testing technology has a corresponding relation to the testing image label; (iii) the LPRST is the first open technical performance evaluation of commercial systems, which identify the strengths and weakness of the existing systems, and the future areas of research in the field of LPR.

Inspec keywords: protocols; image recognition; traffic engineering computing; visual databases

Other keywords: testing technology; labelled license plate image database; evaluation methodology; LPR system test; technical performance evaluation; technical evaluation protocol; structured license plate image database; comprehensive database; commercial systems; license plate recognition systems; comprehensive license plate image database; performance benchmark; image label testing

Subjects: Computer vision and image processing techniques; Image recognition; Traffic engineering computing; Spatial and pictorial databases

http://iet.metastore.ingenta.com/content/journals/10.1049/iet-its.2017.0138
Loading

Related content

content/journals/10.1049/iet-its.2017.0138
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Loading