一個可重組二值影像處理器及其開發平台設計;Design of A Reconfigurable Binary Image Processor and Its Development Platform

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/75941

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/75941

Title:	一個可重組二值影像處理器及其開發平台設計;Design of A Reconfigurable Binary Image Processor and Its Development Platform
Authors:	張瑞慶;Chang, Jui-Ching
Contributors:	資訊工程學系
Keywords:	可重組;二值影像處理器;數學形態學;基因演算法;Real time;Reconfigurable;Binary image processor;Mathematical morphology;Genetic algorithm;Real time
Date:	2018-03-05
Issue Date:	2018-04-13 11:23:12 (UTC+8)
Publisher:	國立中央大學
Abstract:	二值影像處理被運用在多種靜態及串流影像當中，如物件辨識、追蹤、除雜訊，為了適應各種領域不同應用，以及即時影像處理的需求，本研究提出一可重組二值影像處理器以及開發平台，具有高度靈活性與即時影像處理之優點，能夠透過程式控制改變影像處理流程，處理器使用pipeline架構設計，提供形態學與邏輯運算元以及連通元件標記演算法，可重組的特點包含影像大小、二值化門檻值、結構元素及二值影像處理流程。本研究透過合成於Stratix V FPGA平台來驗證硬體效能，其中有16顆運算核心，處理器頻率可達379.94MHz，一秒鐘能夠處理362張大小為"1024×1024" 的影像，並整合基因演算法及簡易開發平台，讓應用系統可以快速的開發，實驗結果證明本系統達到即時二值影像處理，能因應未來各領域之需求。;Binary image processing has considerable applications in image and video processing such as object recognition, tracking, and denoising. To enable real-time image processing for various applications, we propose a reconfigurable binary image processor with a flexible development platform and real-time processing capabilities. The developed processor enables the modification of image processing procedures by reprogramming the employed hardware registers. The processor involves a pipeline architecture design which can perform morphological and binary operations and connected-component labeling. The reconfigurable features of the processor include the image size, image thresholding, structure element, and binary image processing procedure. For performance validation, the processor was implemented on a Stratix V field-programmable gate array, which can process approximately 362 frames per second and can conduct 16 operations with a 1024 × 1024 image at 379.94 MHz. Moreover, proposed platform with a genetic algorithm can be utilized for rapidly and conveniently developing the application system. The experimental results demonstrated that the processor is suitable for use in real-time applications of binary image processing.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	283	View/Open

社群 sharing

Loading...