A framework of reading timestamps for surveillance video
This paper presents a framework to automatically read timestamps for surveillance video. Reading timestamps from surveillance video is difficult due to the challenges such as color variety, font diversity, noise, and low resolution. The proposed algorithm overcomes these challenges by using the deep...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
Samara National Research University
2019-02-01
|
Series: | Компьютерная оптика |
Subjects: | |
Online Access: | http://computeroptics.ru/KO/PDF/KO43-1/430108.pdf |
id |
doaj-af0764fe9c7f434d9d5beed2bcb45a26 |
---|---|
record_format |
Article |
spelling |
doaj-af0764fe9c7f434d9d5beed2bcb45a262020-11-24T21:34:39ZengSamara National Research UniversityКомпьютерная оптика0134-24522412-61792019-02-01431727710.18287/2412-6179-2019-43-1-72-77A framework of reading timestamps for surveillance videoJun Cheng0Wei Dai1Computer School, Hubei Polytechnic University, Huangshi, Hubei, ChinaSchool of Economics and Management, Hubei Polytechnic University, Huangshi, Hubei, ChinaThis paper presents a framework to automatically read timestamps for surveillance video. Reading timestamps from surveillance video is difficult due to the challenges such as color variety, font diversity, noise, and low resolution. The proposed algorithm overcomes these challenges by using the deep learning framework. The framework has included: training of both timestamp localization and recognition in a single end-to-end pass, the structure of the recognition CNN and the geometry of its input layer that preserves the aspect of the timestamps and adapts its resolution to the data. The proposed method achieves state-of-the-art accuracy in the end-to-end timestamps recognition on our datasets, whilst being an order of magnitude faster than competing methods. The framework can be improved the market competitiveness of panoramic video surveillance products.http://computeroptics.ru/KO/PDF/KO43-1/430108.pdfsurveillance videotimestamp localizationtimestamp recognition. |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Jun Cheng Wei Dai |
spellingShingle |
Jun Cheng Wei Dai A framework of reading timestamps for surveillance video Компьютерная оптика surveillance video timestamp localization timestamp recognition. |
author_facet |
Jun Cheng Wei Dai |
author_sort |
Jun Cheng |
title |
A framework of reading timestamps for surveillance video |
title_short |
A framework of reading timestamps for surveillance video |
title_full |
A framework of reading timestamps for surveillance video |
title_fullStr |
A framework of reading timestamps for surveillance video |
title_full_unstemmed |
A framework of reading timestamps for surveillance video |
title_sort |
framework of reading timestamps for surveillance video |
publisher |
Samara National Research University |
series |
Компьютерная оптика |
issn |
0134-2452 2412-6179 |
publishDate |
2019-02-01 |
description |
This paper presents a framework to automatically read timestamps for surveillance video. Reading timestamps from surveillance video is difficult due to the challenges such as color variety, font diversity, noise, and low resolution. The proposed algorithm overcomes these challenges by using the deep learning framework. The framework has included: training of both timestamp localization and recognition in a single end-to-end pass, the structure of the recognition CNN and the geometry of its input layer that preserves the aspect of the timestamps and adapts its resolution to the data. The proposed method achieves state-of-the-art accuracy in the end-to-end timestamps recognition on our datasets, whilst being an order of magnitude faster than competing methods. The framework can be improved the market competitiveness of panoramic video surveillance products. |
topic |
surveillance video timestamp localization timestamp recognition. |
url |
http://computeroptics.ru/KO/PDF/KO43-1/430108.pdf |
work_keys_str_mv |
AT juncheng aframeworkofreadingtimestampsforsurveillancevideo AT weidai aframeworkofreadingtimestampsforsurveillancevideo AT juncheng frameworkofreadingtimestampsforsurveillancevideo AT weidai frameworkofreadingtimestampsforsurveillancevideo |
_version_ |
1725948324017602560 |