Pouzdana detekcija vinove loze korištenjem RGBD kamere

undergraduate thesis

• Title (english): Reliable detection of vines using an RGBD camera
• Author: Petar Slaviček
• Mentor: Zdenko Kovačić (mentor)
• Committee member: Zdenko Kovačić (predsjednik povjerenstva)
• Committee member: Ivo Vatavuk (član povjerenstva)
• Committee member: Ivan Hrabar (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-09, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj rad implementira metodu semantičke segmentacije za detekciju grožđa u vinogradu predstavljenu u (Perez, Rodrigo, 2017). Za razliku od današnjih standarda umjetnih neuronskih mreža i dubokog učenja, ova metoda problemu pristupa konvencionalnijim metodama računalnog vida. Koristi se FRST algoritam za detekciju radijalno simetričnih točaka iz kojih se računaju vektori značajki HOG i LBP deskriptorima. Zatim se SVM klasifikatorom određuje koje točke pripadaju skupu grožđa te se grupiraju DBSCAN algoritmom. Rezultat metode je binarna maska koja označava područja fotografije u kojima je prisutno grožđe. Cilj je korištenje detekcije grožđa u svrhu smanjivanja izgubljenog materijala tijekom prskanja plodova u vinogradu, no ovaj rad samo demonstrira funkcionalnost.

Abstract (english)

This work implements a method of semantic segmentation for grape detection in the vineyard shown in (Perez, Rodrigo, 2017). Unlike today's standards of computer vision using artificial neural networks and deep learning, this method uses more conventional computer vision methods. The FRST algorithm for detecting points of radial symmetry is used from which feature vectors of HOG and LBP descriptors are computed. Then, the SVM classifier determines which points belong to a cluster of grapes and are grouped by the DBSCAN algorithm. The result of the method is a binary mask that indicates areas of images in which grapes are present. The aim is to use grape detection for the purposes of reducing waste while spraying in viticulture, but this work only demonstrates functionality of the method.

• Keywords: Računalni vid; Strojno učenje; Semantička segmentacija; SVM; HOG; LBP; DBSCAN
• Keywords (english): Computer vision; Machine learning; Semantic segmentation; SVM; HOG; LBP; DBSCAN
• Language: croatian
• URN:NBN: urn:nbn:hr:168:755940
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-14 21:39:04