Please use this identifier to cite or link to this item: http://dspace.unimap.edu.my:80/xmlui/handle/123456789/7155
Title: Detection of coconuts location using separability filter and Hough Transform
Authors: Haniza, Yazid
Mohamed Rizon, Mohamed Juhari
Sugisaka, Masanori
Puteh, Saad
Ali Yeon, Md Shakaff
Abdul Rahman, Mohd. Saad
Keywords: Edge detection
Circular Hough Transform
Separability filter
Object detection
Image processing
Computer vision
Hough transforms
Issue Date: 14-May-2005
Publisher: Kolej Universiti Kejuruteraan Utara Malaysia
Citation: p.57-60
Series/Report no.: Proceedings of the 1st International Workshop on Artificial Life and Robotics
Abstract: In this paper, we develop a technique to determine coconuts location using separability filter and Hough transform method. Our research involves two sequential steps for solving the coconuts detection and recognition. The first step uses a modified version of the Hough transform method to detect the region of the image that contains a coconut. The Hough transform has been widely used to extract analytical features such as circles, lines and ellipses. The Hough transform method and several modified versions have been known as robust techniques for curve detection. Before the Hough Transform can be applied, we need edge detection application to identify significant edge point. The next step uses separability filter that had been proposed by Fukui et al.[8 ] to determine the best coconut candidates.
Description: Organized by Kolej Universiti Kejuruteraan Utara Malaysia (KUKUM), 14th - 15th May 2005 at Putra Palace Hotel, Kangar, Perlis.
URI: http://dspace.unimap.edu.my/123456789/7155
Appears in Collections:Conference Papers
Ali Yeon Md Shakaff, Dato' Prof. Dr.
Mohd. Rizon Mohamed Juhari, Prof. Ir. Dr.
Haniza Yazid, Dr.

Files in This Item:
File Description SizeFormat 
Detection of Coconuts Location using Separability Filter.pdfAccess is limited to UniMAP community.452.9 kBAdobe PDFView/Open


Items in UniMAP Library Digital Repository are protected by copyright, with all rights reserved, unless otherwise indicated.