U.S. flag

An official website of the United States government, Department of Justice.

NCJRS Virtual Library

The Virtual Library houses over 235,000 criminal justice resources, including all known OJP works.
Click here to search the NCJRS Virtual Library

Deciphering Severely Degraded License Plates

NCJ Number
308012
Journal
Electronic Imaging Volume: 29 Issue: art00020 Dated: 2017
Author(s)
Shruti Agarwal; Du Tran; Lorenzo Torresani; Hany Farid
Date Published
2017
Length
6 pages
Annotation

This article discusses the presence of useful information in highly degraded images and that convolutional neural networks can be trained to decipher the contents of very low-quality images, specifically, of license plates.

Abstract

Extremely low-quality images, on the order of 20 pixels in width, appear with frustrating frequency in many forensic investigations. Even advanced de-noising and super-resolution technologies are unable to extract useful information from such low-quality images. However, in this paper, the authors show that useful information is present even in such highly degraded images. The authors also show that convolutional neural networks can be trained to decipher the contents of highly degraded images of license plates, and that these networks significantly outperform human observers. Publisher Abstract Provided