Automated number plate recognition from low quality video-sequences
Název práce v češtině: | Automatické rozpoznávání registračních značek aut z málo kvalitních videosekvencí |
---|---|
Název v anglickém jazyce: | Automated number plate recognition from low quality video-sequences |
Klíčová slova: | automatické rozpoznávání značek, konvoluční neuronové sítě, obrazová superrezoluce |
Klíčová slova anglicky: | automated number plate recognition, convolutional neural networks, image super-resolution |
Akademický rok vypsání: | 2017/2018 |
Typ práce: | diplomová práce |
Jazyk práce: | angličtina |
Ústav: | Katedra teoretické informatiky a matematické logiky (32-KTIML) |
Vedoucí / školitel: | Vojtěch Franc |
Řešitel: | skrytý - zadáno a potvrzeno stud. odd. |
Datum přihlášení: | 16.11.2017 |
Datum zadání: | 19.03.2018 |
Datum potvrzení stud. oddělením: | 27.03.2018 |
Datum a čas obhajoby: | 14.06.2018 09:00 |
Datum odevzdání elektronické podoby: | 11.05.2018 |
Datum odevzdání tištěné podoby: | 11.05.2018 |
Datum proběhlé obhajoby: | 14.06.2018 |
Oponenti: | doc. RNDr. Elena Šikudová, Ph.D. |
Zásady pro vypracování |
The existing automated number plate recognition (ANPR) systems usually require still images of a good quality captured by dedicated industrial cameras. The goal of this thesis is to develop ANPR system capable of processing video sequences of a low quality. The low quality videos, captured e.g. by surveillance cameras or mobile phones, are characterized by a low resolution of video frames, motion blur, low dynamic range of luminosity etc. ANPR from low quality videos is required e.g. in forensic applications.
Particular goals for the thesis involve: - Creation of a benchmark for measuring performance of ANPR systems on low quality videos. This will involve design of image processing tools necessary for collection of data with ground truth annotation. - Evaluation of performance of baseline approaches: i) existing commercial ANPR system trained on high quality still images, ii) a human operator. - Design of ANPR system dedicated for processing of low quality videos that will be based on the Convolution Neural Networks learned from examples. |
Seznam odborné literatury |
Textbooks:
- M. Sonka et al. Image Processing, Analysis, and Machine Vision. Chapman and Hall Computing. 1993. - M.I. Schlesinger et al. Ten Lectures on Statistical and Structural Pattern Recognition. Kluwer Academic Publishers. 2002. - I. Goodfellow et al. Deep Learning. MIT Press. 2016. Papers: - S. Du et al. Automatic license plate recognition (alpr): A state-of-the art review. IEEE Trans. Circuits Syst. Video Technol., vol. 23, no. 2, pp. 311–325, 2013. - H. Liu et al. Reading Car License Plates Using Deep Convolutional Networks and LTSMs. ARXIV 2016. - C. Dong et al. Image Super-Resolution Using Deep Convolutional Networks. IEEE Trans. PAMI, vol 38 no 2, 2016. - A. Kappeler et al. Video Super-Resolution With Convolutional Neural Networks. IEEE Trans. Computational Imaging. vol 2, no 2, 2016. |