Please use this identifier to cite or link to this item: https://ptsldigital.ukm.my/jspui/handle/123456789/578205
Title: Neurocomputing approach for firearm identification
Authors: Choong-Yeun Liong (UKM)
Abdul Aziz Jemain (UKM)
Keywords: Firearm classification
Combined images
Geometric moments
Backpropagation neural network
Issue Date: Jan-2018
Description: This paper is an attempt to perceive and order guns using a two-layer neural system model taking into account a feedforward backpropagation calculation. Numerical properties from the joined pictures were utilised for enhanced gun characterisation execution. Inputs of the system model were 747 pictures blackmailed from the discharging pin impression of five differing guns model, Parabellum Vector SPI 9mm. Components created from the dataset were further grouped into preparation set (523 components), testing set (112 components) and acceptance set (112 components). Under managed learning, exact results exhibited that a two-layer BPNN of 11-11-5 arrangement, with tansig/purelin exchange capacities and a “trainlm” preparing calculation, had productively delivered 87% right aftereffect of grouping. The order result serves to be progressed and contrasted with the previous works. Finally, the joined picture districts can offer some accommodating data on the grouping of gun.
News Source: Pertanika Journals
ISSN: 0128-7680
Volume: 26
Pages: 341-352
Publisher: Universiti Putra Malaysia Press
Appears in Collections:Journal Content Pages/ Kandungan Halaman Jurnal

Files in This Item:
File Description SizeFormat 
ukmvital_113386+Source01+Source010.PDF532.2 kBAdobe PDFThumbnail
View/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.