影像處理於渦扇發動機轉子葉片損傷之檢測

NCUIR > Engineering College > Executive Master of Mechanical Engineering > Electronic Thesis & Dissertation > Item 987654321/84299

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/84299

Title:	影像處理於渦扇發動機轉子葉片損傷之檢測
Authors:	李益誠;Lee, Yi-Chen
Contributors:	機械工程學系在職專班
Keywords:	渦扇發動機;壓縮機;葉片;MATLAB;影像處理;凸殼演算法;灰階共生矩陣;turbofan engine;compressor;blade;MATLAB;image process;convex hull;grayscale co-occurrence matrix
Date:	2020-08-13
Issue Date:	2020-09-02 18:57:51 (UTC+8)
Publisher:	國立中央大學
Abstract:	發動機是航空器動力的來源，組成零件眾多，其中葉片可說是發動機中數量最多的一種零組件，其損傷檢查方式主要仍採用人工目視檢查。機器視覺與影像處理常應用於檢測方面，可作為幾何量測、辨識、檢測等。本研究使用CCD攝影機拍攝擷取葉片的圖像，以MATLAB軟體設計操作介面，透過灰階、二值化等影像處理分離影像，利用凸殼演算法與灰階共生矩陣，提出損傷尺寸與表面狀況特徵參數，發展一套檢測系統可以自動檢查判定葉片種類，並能檢測出葉片輪廓損傷，有助於提升檢測精準度，提升檢測效率，減少人力成本，將來可作為自動化檢測的基礎。 ;The engine is the source of aircraft power, and there are many components. Among them, the blade can be said to be the largest number of components in the engine. The visual inspection is still the mainly method of damage inspection. Machine vision and image processing are often used in inspection, and can be used for geometric measurement, identification, and inspection. In this study, a CCD camera was used to capture the image of the high pressure compressor rotor blades, the operation interface was designed by MATLAB software, and the image was separated through image processing such as grayscale and binarization. The convex hull algorithm and the grayscale co-occurrence matrix were used to propose the damage size and surface condition. The development of a detection system can automatically check and determine the types of blades, and can detect blade contour damage, which will help improve detection accuracy, improve detection efficiency, and reduce labor costs. It can be used as the basis for automated detection in the future.
Appears in Collections:	[Executive Master of Mechanical Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	288	View/Open

社群 sharing

Loading...