Graduate Thesis Or Dissertation

 

Image pre-processing algorithms for isolation of defects in Douglas-fir veneer Público Deposited

Conteúdo disponível para baixar

Baixar PDF
https://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/td96k680h

Descriptions

Attribute NameValues
Creator
Abstract
  • The goal of this research is to develop automatic techniques for identifying defects in Douglas-fir veneer. To achieve this goal, computer vision techniques, in particular image-processing, will be applied. Several pre-processing steps may be required to enhance images in making further analysis easier, faster, and more reliable. Pre-processing is often necessary when analyzing natural materials due to the amount of complex variation present. However, if such analysis is associated with several compute-intensive passes over a large data set, the possibilities of what can be achieved, is very limited. It is advantageous to reduce the data in a way that little or no information is lost that may affect the image analysis. This reduced image is then used in subsequent analysis to determine the actual type of defect, its extent and exact location. Whether such a reduction is possible usually depends on the intended application. For Douglas-fir veneer sheets at the plugging stage, the material is relatively clear, with defects occupying only a minor portion of the surface. It will be sham that reduction by pre-processing is indeed feasible. This study investigates three potential image sweep and mark (ISM) pre-processirg algorithms that perform this kind of reduction. These algorithms are very diverse in their methods of operation and include methodologies based on first-order statistics, mathematical morphology, and color cluster analysis. It is shown in the conclusions of this thesis that one of these algorithms is capable of reducing the data of a typical Douglas-fir veneer image by 80% with an error rate of only 0.3%. It is also shown that color is a significant factor to be considered for the detection of defects.
Resource Type
Date Available
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Non-Academic Affiliation
Subject
Declaração de direitos
Publisher
Language
Digitization Specifications
  • File scanned at 300 ppi (Monochrome, 24-bit Color) using ScandAll PRO 1.8.1 on a Fi-6670 in PDF format. CVista PdfCompressor 4.0 was used for pdf compression and textual OCR.
Replaces

Relações

Parents:

This work has no parents.

Em Collection:

Itens