KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Settings

Author: search.filters.author.Kanjanasurat, IsoonSubject: search.filters.subject.Vessel extraction

Search Results

Now showing 1 - 2 of 2
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Retinal Blood Vessel Extraction by Using Pre-processing and IterNet Model
    (2020-12-03)
    Tenghongsakul, Kasi
    ;
    Kanjanasurat, Isoon
    ;
    Purahong, Boonchana
    ;
    Lasakul, Attasit
    At present, many of visual disease happened from the abnormality of retinal vessels. The automatic vascular extraction from fundus images is essential for the diagnosis to reduce vision loss. This paper offers retinal blood vessel segmentation using the pre-processing and IterNet model, a convolution neural network. The green channel and gray scale image that is high contrast between the blood vessel and background, including the normalization, were used to improve blood vessel image quality. The proposed method was tested with two widely used databases, including DRIVE and CHASEDB-1, which unique characteristics in each data set. The results of blood vessel extraction of Drive and CHASEDB-1 achieved sensitivity 0.8126 and 0.7541, respectively.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Personal identification using a delaunay triangle and optic disc retinal vascular pattern
    (2020-01-01)
    Kanjanasurat, Isoon
    ;
    Purahong, Boonchana
    ;
    Aoyama, Hisayuki
    ;
    Benjangkaprasert, Chawalit
    ;
    Pintavirooj, Chuchart
    Retinal vascular patterns are unique and individual. They provide highly secure and correct identity authentication. In this study, we exploit an image alignment approach based on a geometric invariant, which is the area spanned by feature-point triplets for personal identification. First, we located the optic disc by using a projection of the vascular structure in vascular extraction and extracted feature points that are bifurca- tions of a retinal blood vessel in the vicinity of the optic disc as the landmarks. Delaunay triangulation is then applied to the extracted feature points. The absolute invariant is then derived by taking the ratio of successive triangular area patches. The alignment is achieved by establishing correspondences between feature points after a conformal sort- ing step based on a derived set of absolute affine invariants. The affine transformation parameters can then be calculated by the corresponding vertices of the most robust neigh- bouring triangle of both inquiry and reference images. The optic disc localization results successfully located 95.95% in six widely used retinal image databases. The algorithm of vascular extraction, applied on the DRIVE database, provided an average accuracy of ap- proximately 94.1%. The best accuracy and sensitivity for neighbouring triangle matching obtained were 99.90% and 87.66%, respectively.