(For USM Staff/Student Only)

EngLib USM > Ω School of Electrical & Electronic Engineering >

Lossless karhunen-loève transform - embedded approach (Raspberry Pi)

Lossless karhunen-loève transform - embedded approach (Raspberry Pi) / Muhammad Mohd Salleh
Kajian ini membentangkan Ketiadaan Kehilangan dalam Karhunen-Loève Transform (KLT) Sistem Terbenam. Kajian ini akan diselarikan menggunakan Open Multi Processing alam sekitar (OpenMP) ke dalam Model Raspberry Pi 3 B. Integer KLT dipilih kerana menunjukkan yang prestasi terbaik dalam mengurai komponen spektrum untuk proses pemampatan imej hyperspectral. Ini kerana masih kurang kajian dibuat pada platform terbenam disebabkan oleh kerumitan untuk diaplikasi. Persekitaran OpenMP digunakan untuk menyelarikan proses pemampatan imej ke dalam sistem multicore itu. Algoritma akan digunakan ke dalam Raspberry Pi, yang mempunyai 4-teras. Cara untuk memulakaan penggunaan papan Raspberry Pi, Linux (Raspbian) digunakan sebagai sistem operasinya yang mana boleh di dapati di laman sesawang Raspberry Pi. Perisian Geany digunakan untuk melaksanakan algoritma Integer KLT dalam kod C. Prestasi algoritma diukur melalui masa pelaksanaan menggunakan beberapa tahap pengelompokan pada AVIRIS imej hyperspectral. Tesis ini berjaya apabila imej hyperspectral AVIRIS dapat digunakan ke dalam Raspberry Pi dalam julat 920.724s untuk 28.482s untuk pelbagai kelompok 1-56 kelompok. Hasil daripada pelaksanaan dibandingkan dengan pendekatan platform perbezaan; desktop dan sistem Pemprosesan Isyarat Digital (DSP). _______________________________________________________________________________________________________ The research presented the Lossless Karhunen-Loève Transform (KLT)-Embedded System in parallelize using Open Multi Processing (OpenMP) environment using the Raspberry Pi 3 Model B. The Integer KLT is chosen because its show the superior performance in decorrelating the spectral component in hyperspectral image compression. This is due to the less implementation attempt has yet been made on an embedded platform due to the complexity issues. The OpenMP environment is used to parallelize the image compression process into the multicore architecture. The algorithm will be deployed into the Raspberry Pi, which have 4-cores processing capability. In order to initialize the Raspberry Pi, Linux (Raspbian) is used as its operating system. Next, Geany software is used to implement the Integer KLT algorithm in C language. The performance of the algorithm is measured through the execution time using several levels of clustering of on AVIRIS hyperspectral image. This thesis is able to execute the AVIRIS hyperspectral image using Raspberry Pi within range 920.724s to 28.482s for clustering range 1-56 cluster. The outcome from the implementation is compared with difference platform approaches; desktop and Digital Signal Processing (DSP) system.
Contributor(s):
Muhammad Mohd Salleh - Author
Primary Item Type:
Final Year Project
Identifiers:
Accession Number : 875007225
Barcode : 00003107104
Language:
English
Subject Keywords:
Karhunen-Loève Transform; Open Multi Processing; hyperspectral image compression
First presented to the public:
6/1/2017
Original Publication Date:
4/17/2018
Previously Published By:
Universiti Sains Malaysia
Place Of Publication:
School of Electrical & Electronic Engineering
Citation:
Extents:
Number of Pages - 57
License Grantor / Date Granted:
  / ( View License )
Date Deposited
2018-04-17 10:00:35.082
Date Last Updated
2019-01-07 11:24:32.9118
Submitter:
Mohd Jasnizam Mohd Salleh

All Versions

Thumbnail Name Version Created Date
Lossless karhunen-loève transform - embedded approach (Raspberry Pi)1 2018-04-17 10:00:35.082
Engineering Library USM @ 2017