Određivanje biomase i veličine ribe u morskom kavezu na temelju skupa podataka iz podvodnog sonara

master's thesis

• Title (english): Determination of biomass and fish size in a marine cage based on a set of data from underwater sonar
• Author: Kristijan Matan Čagalj
• Mentor: Zdenko Kovačić (mentor)
• Committee member: Zdenko Kovačić (član povjerenstva)
• Committee member: Tamara Petrović (član povjerenstva)
• Committee member: Goran Vasiljević (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-14, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

U sklopu ovoga rada je kreiran program koji služi generiranju slika potrebnih za kreiranje neuronske mreže. Tako kreirana neuronska mreža se može koristiti za određivanje ukupne biomase riba u kavezu riblje farme. Sam program se temelji na obradi podataka dobivenih iz sonara Aris 3000 koji dolaze u obliku videa. Video se obrađuje tako da se za početnu obradu koristi ARISCOPE aplikacija te se potom iz videa dobivaju frame-ovi od kojih se svaki pojedinačno obrađuje prolazima kroz potrebne filtere te se u konačnici dobivaju dva skupa slika. Jedan skup slika se koristi za određivanje ukupne biomase riba unutar kaveza dok se iz drugog skupa slika određuje histogram mase pojedinačnih riba koji se potom i grafički prikazuje.

Abstract (english)

As part of this paper, a program was created that serves to generate the images needed to create a neural network. The neural network that is created can be used to determine the total fish biomass in a fish farm cage. The program itself is based on the processing of data obtained from the Aris 3000 sonar that comes in the form of video. The video is processed in such a way that the ARISCOPE application is used for the initial processing, and after that, frames are obtained from the video, each of obtained image is processed individually by passing through the necessary filters. At the end, two sets of images are obtained. One set of images is used to determine the total biomass of fish inside the cage, while the other set of images determines the histogram of the mass of individual fish, which is then graphically displayed.

• Keywords: Estimacija biomase; masa riba; histogram mase riba; skup slika za učenje neuronske mreže; umjetna inteligencija; strojno učenje
• Keywords (english): Biomass estimation; fish mass; fish mass histogram; neural network learning dataset; artificial intelligence; machine learning
• Language: croatian
• URN:NBN: urn:nbn:hr:168:767186
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije (magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 09:52:34