Scopus
Novel approaches to determine age and gender from dental x-ray images by using multiplayer perceptron neural networks and image processing techniques
Chaos, Solitons and Fractals · Mart 2019
Makale Bilgileri
DergiChaos, Solitons and Fractals
Yayın TarihiMart 2019
Cilt / Sayfa120 · 127-138
Scopus ID2-s2.0-85060920184
Özet
It may be necessary to determine the identity or gender of a person for any reason (disasters, inheritance etc.). In such cases, forensic medical institutions are asked for help. Forensic science institutions try to estimate the age of people's teeth and bones. In this study, a novel algorithm was developed to keep these predictions at the highest level and to obtain definite results. The data base of 162 different tooth classes is created manually. All image sizes are 150×150 pixels. First, image preprocessing techniques have been applied to teeth images. These preprocessing techniques were first applied to teeth images. After this process, the segmentation process of the teeth images was performed to extract the feature by novel segmentation algorithm. Segmentation can be done automatically and dynamically. Numerical data obtained as a result of feature extraction from dental images is presented as an inputs to Multi layer perceptron neural network. In application, feature reduction can be performed. Thanks to the originally developed algorithm, the highest success rates were obtained with the highest 99.9% (full segment) and 100% (notfull segment) classification. After classification, for many dental groups the age estimate is performed with zero error. Application was developed as a multidisciplinary study.
Yazarlar (2)
1
Emre Avuçlu
2
Fatih Başçiftçi
Anahtar Kelimeler
Age estimation
Backpropagation algorithm
Image processing techniques
Multilayer perceptron
Panoramic x-ray
Kurumlar
Aksaray Üniversitesi
Aksaray Turkey
Selçuk Üniversitesi
Selçuklu Turkey
Metrikler
28
Atıf
2
Yazar
5
Anahtar Kelime