OPG Images Preprocessing Enhancement for Diagnosis Purposes

Orthopantomogram (OPG) image is considered one of the techniques which plays a vital role in the dental diagnosis of diseases and its treatment. Generally speaking, the panoramic OPG images acquired with Orthopantomogram are corrupted by noise and have a poor visualization. For this reason, appropriate techniques for suppressing the noise and contrast improvement are required for the development. The current paper proposed an algorithm which include two-phases, noise reduction and contrast enhancement for OPG images. For noise reduction, adaptive median filter and discrete wavelet transform have been used for handling mixed noises which contains Salt and Pepper noise, Gaussian noise, and Speckle noise. In other words, this phase is capable of suppressing the noise even though preserving the important feature in the images. Subsequently, the contrast of OPG is enhanced by using morphology processing and contrast limited histogram equalization (CLAHE). Furthermore, the performance of noise reduction and contrast enhancement are evaluated separately with visual results and quantitative measurement. Moreover, noise reduction is in comparison averaged PSNR of 31.78 dB and averaged SSIM of 0.9115 for mixed noise in comparison with other filtrations techniques while the contrast enhancement phase is achieved an average AMBE of 27.356 and an average EME of 51.631 Module.