KMITL

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

Browse

Settings

Author: search.filters.author.Satayarak, Nitjaree

Search Results

Now showing 1 - 3 of 3
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Ensemble of CNN classifiers using Choquet Fuzzy Integral Technique for PCB Defect Classification
    (2024-01-01)
    Purahong, Boonchana
    ;
    Krungseanmuang, Woranidtha
    ;
    Tenghongsakul, Kasi
    ;
    Archevapanich, Tuanjai
    ;
    Khunthawiwone, Parkpoom
    This paper presents a novel method for detecting defects in printed circuit boards (PCBs) using an ensemble of classifiers based on the Choquet fuzzy integral. Our approach employs convolutional neural network (CNN) models, specifically ResNet152, VGG19, and InceptionV3 as base classifiers to identify six types of PCB defects: spurs, mouse bites, short circuits, open circuits, spurious copper, and pinholes. Given the critical role of PCBs in ensuring electronic equipment reliability, effective defect detection methods like ours are essential. We employ pre-trained CNN models for feature extraction and classification of PCB defects. Following this, we combine the prediction scores using the Choquet fuzzy integral to derive more accurate final labels, exceeding the accuracy of standalone models. Our approach is tested on PCB images obtained from public repositories, captured using a linear scan CCD. The evaluation results demonstrate average precision, recall, F-score, and accuracy of 93.0%, 95.2%, 95.1%, and 95.1%, respectively.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Optic Disk and Fovea Localization by Using the Direction of Blood Vessels and Morphology Operation
    (2021-03-17)
    Kanjanasurat, Isoon
    ;
    Satayarak, Nitjaree
    This paper presents the optic disk localization by using the matrix that extracted the blood vessels' direction and finding the fovea position using morphology operation in diabetic retinopathy. Our approach begins with blood vessel extraction for locating the optic disk area. Next process, the blood vessel structure was used to estimate the location of the optic disk. Next step, the morphology operator, including erosion and dilation, was used to prepare for attaining the fovea region. Finally, the location of the fovea was estimated by using the position of the optic disk, and specific characteristics of the fovea spot. The proposed method was tested on the DRIVE, DIARETDB0, and DIARETDB1 that is a public diabetic retinal image dataset. The results of the optic disk and fovea localization were compared with the ground truth image. This method can locate optic disk and fovea on DRIVE 100%. In DIARETDB0 and DIARETDB1, this algorithm can achieve optic disk 96.15% and 98.87%, respectively, and locate fovea more than 90%.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Blood Vessel Extraction and Optic Disk Localization for Diabetic Retinopathy
    (2020-09-15)
    Kanjanasurat, Isoon
    ;
    Purahong, Boonchana
    ;
    Pintavirooj, Chuchart
    ;
    Satayarak, Nitjaree
    ;
    Benjangkaprasert, Chawalit
    This paper presents methods of vascular extraction and optic disk localization in the retinal images. Our approach begins with preprocessing to improve the quality of blood vessels. In the next step, a matrix filter was applied to express blood vessels. Finally, the blood vessel structure was used to estimate the location of the optic disk. The proposed method was tested on all different forty retinal images from the DRIVE database, which public retinal image dataset. The results of vessel extraction were compared with the ground truth image. The error of vascular extraction showed that the average sensitivity and accuracy were 79.81% and 94.98%, respectively. The optic disk localization achieved 97.5%.