Title: Restoration of dental radiographs corrupted by quantum noise via a noise-adaptive sub-global weighted sum of radiometrically similar pixels

Authors: Manoj T. Joy; B. Priestly Shan; Geevarghese Titus

Addresses: Computer Science and Engineering, Sathyabama Institute of Science and Technology, Chennai, TN, India ' Electronics and Communication Engineering, Chandigarh University, Mohali, Punjab, India ' Electronics and Communication Engineering, Amal Jyothi College of Engineering, Kanjirappally, India

Abstract: The image creation principle causes quantum noise in dental X-rays to interfere with image processing like registration and segmentation. The detector produces a blurry image without an object. Dental X-rays are dangerous. These factors slow image registration and segmentation. Adaptive total variation regularisation reduces noise. This study presents a customised method for suppressing quantum noise in dental X-rays. In the noise-adaptive sub-global weighted sum of radiometrically similar pixels (NSWSRSP) technique, the denoised pixel value is a weighted sum of radiometrically comparable pixels in a distorted image search-block. Noise reduction is adjustable since the smoothing parameter is proportional to the noise standard deviation in a homogeneous region. The peak-signal-to-noise ratio (PSNR) in dB for total variation (TV), directional weighted median (DWM), bilateral, anisotropic diffusion (AD), wavelet thresholding (WT), block matching (BM), and the proposed technique are 30.7619 ± 1.1744, 22.0260 ± 2.9462, 28.9104 ± 4.0115, 30.4842 ± 4.8967, 33.6781 ± 1.5863, 23.6935 ± 2.5358, and 33.8648 ± 1.6825 respectively. NSWSRSP can extract noise-free dental radiographs from quantum-noise-corrupted observations using the described technique.

Keywords: dental X-ray images; image denoising; preprocessing; quantum noise; restoration.

DOI: 10.1504/IJCBDD.2022.128246

International Journal of Computational Biology and Drug Design, 2022 Vol.15 No.3, pp.249 - 266

Received: 13 Jul 2022
Accepted: 26 Aug 2022
Published online: 12 Jan 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article