Lokalizacija mej med odbojnimi površinami v deflektometrični aplikaciji

diplomsko delo

Janja Koželj (Author), Matej Kristan (Mentor)

Abstract

V diplomski nalogi obravnamo problem detekcije mej med odbojnimi površinami med postopkom skeniranja z deflektometrijo. V tem postopku se na površino projicira črtast vzorec, kar močno otežuje lokalizacijo meje s standardnimi metodami. Zato predlagamo novo metodo za lokalizacijo mej, ki uporablja konvolucijsko nevronsko mrežo in aktivne konture. Uspešnost naše metode demonstriramo na slikah, kjer je prikazan stranski pogled avtomobila, detektiramo pa meje med vrati. Predlagana mreža je oblike enkoder dekoder in za boljše prepoznavanje vzorca uporablja razširjene konvolucijske sloje. Za najboljši postopek prileganja se izkaže robustno prileganje na segmentirani maski z uporabo aktivnih kontur, ki izboljša povprečno napako iz 5.00 na 2.67 pikslov. Predlagana metoda na testni množici doseže natančnost 0.91, priklic 0.68 in F-mero 0.76, procesiramo pa lahko približno 4.29 slik na sekundo.

Keywords

konvolucija;nevronske mreže;semantična segmentacija;aktivne konture;deflektometrija;računalništvo;računalništvo in informatika;računalništvo in matematika;interdisciplinarni študij;univerzitetni študij;diplomske naloge;

Data

Language:	Slovenian
Year of publishing:	2020
Typology:	2.11 - Undergraduate Thesis
Organization:	UL FRI - Faculty of Computer and Information Science
Publisher:	[J. Koželj]
UDC:	004.8(043.2)
COBISS:	29404675
Views:	687
Downloads:	107
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Border localization between reflective surfaces in a deflectometry application
Secondary abstract:	In this thesis, we address the problem of border detection between reflective surfaces during the process of deflectometry. In this process a striped pattern is projected on the surface, which makes it difficult to localize border with standard methods. To address this problem, we propose a new method for border localization, which uses convolutional neural network and active contours. We demonstrate the performance of our method on the task of car door border detection from images taken from the side of a car. Proposed network has the encoder decoder architecture and contains dilated convolutional layers for better pattern recognition. We show that robust fitting on segmented masks using active contours is the best way of fitting, and it reduces mean error from 5.00 to 2.67 pixels. On test set the proposed method achieves precision of 0.91, recall of 0.68 and F-score of 0.76. The method allows processing at approximately 4.29 frames per second.
Secondary keywords:	convolution;neural networks;semantic segmentation;active contours;deflectometry;computer science;computer and information science;computer science and mathematics;interdisciplinary studies;diploma;
Type (COBISS):	Bachelor thesis/paper
Study programme:	1000407
Embargo end date (OpenAIRE):	1970-01-01
Thesis comment:	Univ. v Ljubljani, Fak. za računalništvo in informatiko
Pages:	58 str.
ID:	12031215