Automatizirano prepoznavanje biljaka na fotografijama koje su snimane u prirodnom staništu

master's thesis

• Title (english): Automated recognition of plants in photographs taken in their natural habitat
• Author: Ivan Baljkas
• Mentor: Mirko Randić (mentor)
• Committee member: Mirko Randić (predsjednik povjerenstva)
• Committee member: Bruno Blašković (član povjerenstva)
• Committee member: Luka Humski (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-11, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Aktualno pitanje jest očuvanje ugroženih biljnih vrsta i biljnih vrsta koje imaju važnu ulogu u eko-sustavu. S obzirom na to da je teško pronaći veći broj stručnjaka koji mogu raspoznavati i prebrojavati određene biljne vrste, praktično bi bilo implementirati automatiziranu detekciju cvjetova na slikama. Imajući na umu da svatko pri ruci ima kameru, a postoji i mogućnost korištenja dronova, vrlo lako se može snimiti velika površina eko-sustava. Dobiveni podaci u obliku slika će se najprije koristiti za pripremu sustava za detekciju koji će nakon toga moći automatizirano detektirati cvjetove s novih slika. Za implementaciju takvog sustava potrebno je znanje iz umjetne inteligencije i strojnog učenja. Uži dio strojnog učenja koji se bavi detekcijom slika zove se duboko učenje, a model specijaliziran za takve probleme jest konvolucijska neuronska mreža. Također, uvođenjem YOLO algoritma u implementaciju donosi nam poželjne karakteristike sustava za detekciju koji smo razvili i opisali u okviru ovog rada.

Abstract (english)

The current issue is the conservation of endangered plant species and plant species that play an important role in the ecosystem. Since it is difficult to find a large number of experts who can recognize and count certain plant species, it would be practical to implement automated flower detection in images. Keeping in mind that everyone has a camera at hand, and there is also the possibility of using drones, it is very easy to capture a large area of the ecosystem. The obtained data in the form of images will first be used to prepare a detection system that will then be able to automatically detect flowers from new images. The implementation of such a system requires knowledge of artificial intelligence and machine learning. The narrower part of machine learning that deals with image detection is called deep learning, and the model specializing in such problems is the convolutional neural network. Also, by introducing the YOLO algorithm into implementation, it brings us the desirable characteristics of the detection system that we have developed and described in this paper.

• Keywords: Umjetna inteligencija; strojno učenje; duboko učenje; neuronska mreža; konvolucijski model; TensorFlow; Keras; OpenCV; detekcija slika; YOLO algoritam; cvijet
• Keywords (english): Artificial Intelligence; machine learning; deep learning; neural network; convolutional model; TensorFlow; Keras; OpenCV; image detection; YOLO algorithm; flower
• Language: croatian
• URN:NBN: urn:nbn:hr:168:524268
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-14 08:22:37