diplomsko delo
Abstract
V diplomskem delu smo se ukvarjali z razvojem preprostega programskega sistema za
zaznavo gibanja na osnovi računalniškega vida, ki se izvaja na usmerjevalniku. Sistem
sestoji iz usmerjevalnika, na katerem je naložen operacijski sistem OpenWrt, spletne
kamere in treh različnih računalniških algoritmov. Za zaznavanje gibanja smo
implementirali naslednje tri algoritme, in sicer algoritem slika razlik, algoritem modeliranje
ozadja in algoritem akumulirana slika razlik. Razvite algoritme smo pod istimi pogoji (tj. ob
istem številu zajetih slik in isti sceni) preizkusili tako na osebnem računalniku, kakor tudi
na usmerjevalniku. Rezultati so pokazali, da se zaradi svoje preprostosti najhitreje izvede
algoritem slika razlik. Ugotovili pa smo tudi, da zaznavanje gibanja s tem algoritmom ni
tako zanesljivo kot pri ostalih dveh algoritmih, ki sicer obdelata manjše število slik na
sekundo. S pomočjo algoritma slika razlik lahko v povprečju obdelamo 34 slik na sekundo,
z najpočasnejšim algoritmom, tj. algoritem modeliranje ozadja, pa obdelamo 9 slik na
sekundo, pri čemer smo uporabili usmerjevalnik ASUS WL-500G V1.
Keywords
usmerjevalniki;kamera;OpenWrt;programski jezik C;računalniški vid;zaznavanje gibanja;obdelava digitalnih slik;
Data
Language: |
Slovenian |
Year of publishing: |
2013 |
Typology: |
2.11 - Undergraduate Thesis |
Organization: |
UM FERI - Faculty of Electrical Engineering and Computer Science |
Publisher: |
[J. Rezec] |
UDC: |
004.43:004.92(043.2) |
COBISS: |
17083926
|
Views: |
1496 |
Downloads: |
129 |
Average score: |
0 (0 votes) |
Metadata: |
|
Other data
Secondary language: |
English |
Secondary title: |
DETECTION OF EXTRAORDINARY EVENTS USING WEBCAM CONNECTED TO A ROUTER |
Secondary abstract: |
In this diploma thesis we were focused on a development of a simple program system.
This system, running on a router, is designed to detect motion based on the computer
vision. It consists of a router supported by OpenWrt operating system, webcam and of
three different motion detection algorithms, which are: frame difference algorithm,
background modeling algorithm and accumulative frame difference algorithm. We tested
these algorithms under the same conditions (i.e., the same scene and frames captured)
on a personal computer and also on router. We found out, that the simplest frame
distance algorithm was the fastest, but not as reliable at motion detection as the other two
algorithms, that process less frames per second. Frame difference algorithm has the
average of processing 34 frames per second, while the slowest background modeling
algorithm has the average of 9 frames per second. All algorithms were tested and
compared running on ASUS WL-500G V1 router. |
Secondary keywords: |
router;OpenWRT;programming language C;computer vision;motion detection;digital image processing; |
URN: |
URN:SI:UM: |
Type (COBISS): |
Bachelor thesis/paper |
Thesis comment: |
Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko |
Pages: |
VII, 45 f. |
ID: |
8726237 |