Disaster Victim Identification: Automatic Dental Radiograph Image Recognition Using Fuzzy Approach

This research proposes an automatic dental radiograph image recognition using fuzzy approach for disaster victim identification. Moreover, a new method using fuzzy inference system to determine an optimal degree of Zernike moment in the system is also proposed. Correlation coefficient and Euclidean distance are used as input variables for fuzzy set. The difference between the correlation coefficients of the Zernike moment of adjoining degrees and the total difference of correlation coefficient are used as fuzzy set for input variable. The difference between Euclidean distance of Zernike moment of adjoining degrees, and the total difference of Euclidean distance are also used as a fuzzy set for input variable. The similarity degree of dental x-ray images is used for output variable. Seven fuzzy rules are used for mapping input variable into output variable. The experiments are conducted on 56 dental x-ray images obtained from some hospitals. The proposed method gets 100% accuracy in matching between original and rotated images, and 98.2% accuracy in matching between original and distorted images. The proposed method is three times faster than conventional method in average execution time. These results indicate that the proposed method can improve searching accuracy and efficiency of the time in the process of disaster victim identification using dental structures.