Panoramic Radiographic Images Processing for Age Detection System Using Fractal Method

The victim from any disaster needs to make some forensic identification. Some corpses that might be broken caused by a bomb, burned, perished, and a mass accident could not be identified using conventional methods such as face and fingerprint. Forensic odontology identifies the victim using their molar teeth because the accuracy is high (same as a fingerprint). The tooth is the stronger part of our body to estimate a person's age from parental to adult. One of the important things needed to identify the victim is age. Tooth pulp can be classified as someone's age because the older the large of pulp is getting smaller. Panoramic radiography images of first mandibular molar teeth from 14 until 60 years old are used to identify using fractal method and K-Nearest Neighbor (K-NN) classification. The first step is acquiring the image, then pre-processing process to separate the pulp from teeth. After that, the image of the pulp is being extracted using fractal future extraction. Then it classifies into four groups using KNN. Image of 376 research data are used for test data. The accusation of this system achieves around 82 % with threshold = 0.8, fractal dimension
$=1, \mathrm{s}$
matrix fractal
$=[2]$
, and
$\mathrm{K}$
value
$=1$
with Euclidean distance.