Comparative Analysis of Non-Local Means and Hybrid Median Bilateral Filtering for Mobile Image Denoising

Abstract

Image denoising is an important step in many computer vision applications, including but not limited to intelligent transportation systems, mobile robotics, remote sensing, autonomous driving, and surveillance. In wearable devices such as smart glasses, wearable health monitoring devices, and other body-mounted devices, the image quality suffers due to non-uniform illumination, motion, occlusions, etc. The image denoising is required prior to image analysis for many computer vision tasks. Many computer vision algorithms perform well when the images are captured using high-resolution cameras under controlled conditions. However, the performance of these methods degrades when images are captured using low-resolution mobile phone cameras or when the images suffer from environmental noise. The images captured using mobile phones act as a proxy for many wearable devices to a reasonable extent because of the similarity in the size of sensors used and the imaging conditions. In this paper, the images captured by a mobile phone are subjected to Gaussian noise. Two different methods, namely, the non-local means denoising method and the hybrid median bilateral method, are used for denoising the images. The performance evaluation of the two methods is done using metrics, namely, peak signal-to-noise ratio and structural similarity index. It is observed that the non-local means denoising method performs better than the hybrid median bilateral denoising method. The method is useful in many image analysis tasks for images captured with smart wearable devices. Received: 30 October 2025 | Revised: 18 March 2026| Accepted: 2 April 2026



Conflicts of Interest

The author declares that he has no conflicts of interest to this work.


Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.



Author Contribution Statement

Ashwani Kumar Aggarwal: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration.