Weld Defect Radiographic Image Segmentation with Finite Mixture Model (FMM)

Authors: , ,

Abstract: Image segmentation is an important task in computer vision. This paper aims at studying two image segmentation methods based on the mixture of two probability density functions. We explore here the exploitation of a Finite Mixture Model (FMM), particularly Gaussian and Student’s t - mixtures models (GMM, SMM) in histogram classification-based image segmentation. The expectation maximization (EM) algorithm is used to estimate the parameters of each model that maximize the log-likelihood function. Experiments have been conducted to segment real industrial radiography images. Comparison of results is achieved for GMM and SMM image segmentation. The obtained results show that the SMM is more robust that the GMM when segmenting real industrial radiography images.

Engineering World, E-ISSN: 2692-5079, Volume 2, 2020, Art. #20

PDF
XML
Cite
×
Citation Tools
WSEAS AMA
Zahia Guezoui, Aicha Baya Goumeidane, Nafaa Nacereddine , "Weld Defect Radiographic Image Segmentation with Finite Mixture Model (FMM)," Engineering World, vol. 2, pp. 134-138, 2020, DOI:
Zahia Guezoui, Aicha Baya Goumeidane, Nafaa Nacereddine . Weld Defect Radiographic Image Segmentation with Finite Mixture Model (FMM). Engineering World. 2020;2:134-138.
Copy to Clipboard
Citation copied to Clipboard

Certification

---