Yazar "Almryad, Ayaad Saad S." seçeneğine göre listele

Listeleniyor 1 - 1 / 1
  • Küçük Resim
    Öğe
    IDENTIFICATION OF BUTTERFLY SPECIES USING MACHINE LEARNING AND IMAGE PROCESSING TECHNIQUES
    (2020-02-28) Almryad, Ayaad Saad S.
    In today’s competitive conditions, producing fast, inexpensive and reliable solutions are objectives for engineers. Development of artificial intelligence and the introduction of this technology to almost all areas have created a need to minimize the human factor by using artificial intelligence in the field of image processing, as well as to make a profit in terms of time and labor. In this thesis, we propose an automated butterfly species identification model using artificial neural and deep neural networks. The study in the thesis consists of two stages. In the first stage, we studied on lab-based butterfly images taken on under a fixed protocol. The species of butterflies in these images are identified by expert entomologists. We used a total of 140 images for lab-based butterfly images of 10 species. After applying some preprocess to the images such as histogram equalization and background removing, we extracted several features from the butterfly images. Finally, we used an artificial neural network in MATLAB version R2014b using the Neural Network Toolbox for butterfly identification. The ANN model achieved an accuracy of 98%. In the second stage of the thesis, we studied on field-based butterfly images. We collected 44659 images of 104 different butterfly species taken with different positions of butterflies, the shooting angle, butterfly distance, occlusion, and background complexity in the field in Turkey. Since many species have a few image samples we constructed a field-based dataset of 17769 butterflies with 10 species. Convolutional Neural Networks (CNNs) implemented by Python were used for the identification of butterfly species. Comparison and evaluation of the experimental results obtained using three different network structures are conducted. Experimental results on 10 common butterfly species showed that our method successfully identified various butterfly species.