An Efficient Road Image Dehazing Model Based on Entropy-Weighted Gaussian Mixture Model and Level Set Refinement for Autonomous Driving Applications
Journal Title: Mechatronics and Intelligent Transportation Systems - Year 2025, Vol 4, Issue 1
Abstract
Foggy road conditions present significant challenges for road monitoring systems and autonomous driving, as conventional defogging techniques often fail to accurately recover fine details of road structures, particularly under dense fog conditions, and may introduce undesirable artifacts. Furthermore, these methods typically lack the ability to dynamically adjust transmission maps, leading to imprecise differentiation between foggy and clear areas. To address these limitations, a novel approach to image dehazing is proposed, which combines an entropy-weighted Gaussian Mixture Model (EW-GMM) with Pythagorean fuzzy aggregation (PFA) and a level set refinement technique. The method enhances the performance of existing models by adaptively adjusting the influence of each Gaussian component based on entropy, with greater emphasis placed on regions exhibiting higher uncertainty, thereby enabling more accurate restoration of foggy images. The EW-GMM is further refined using PFA, which integrates fuzzy membership functions with entropy-based weights to improve the distinction between foggy and clear regions. A level set method is subsequently applied to smooth the transmission map, reducing noise and preserving critical image details. This process is guided by an energy functional that accounts for spatial smoothness, entropy-weighted components, and observed pixel intensities, ensuring a more robust and accurate dehazing effect. Experimental results demonstrate that the proposed model outperforms conventional methods in terms of feature similarity, image quality, and cross-correlation, while significantly reducing execution time. The results highlight the efficiency and robustness of the proposed approach, making it a promising solution for real-time image processing applications, particularly in the context of road monitoring and autonomous driving systems.
Authors and Affiliations
Muhammad Zeeshan Naeem
Keywords
Related Articles
- EP ID EP764543
- DOI https://doi.org/10.56578/mits040102
- Views 10
- Downloads 0
A Comprehensive Exploration of Resource Allocation Strategies within Vehicle Ad-Hoc Networks
In recent years, a surge in the utilisation of vehicle-to-vehicle (V2V) communication has been observed, serving as a pivotal factor in facilitating automatic control of vehicles without human intervention. This advancem...
Enhancing Occluded Pedestrian Re-Identification with the MotionBlur Data Augmentation Module
In the field of pedestrian re-identification (ReID), the challenge of matching occluded pedestrian images with holistic images across different camera views is significant. Traditional approaches have predominantly addre...
A Robust Road Image Defogging Framework Integrating Pythagorean Fuzzy Aggregation, Gaussian Mixture Models, and Level-Set Segmentation
Foggy road conditions present substantial challenges to road monitoring and autonomous driving systems, as existing defogging techniques often fail to accurately recover structural details, manage dense fog, and mitigate...
Optimizing Electric Vehicle Charging Infrastructure: A Site Selection Strategy for Ludhiana, India
This study investigates the spatial distribution and potential expansion of electric vehicle (EV) charging infrastructure in Ludhiana, India, with a focus on optimizing site selection to accommodate increasing demand. A...
An Adaptive Multi-Stage Fuzzy Logic Framework for Accurate Detection and Structural Analysis of Road Cracks
The degradation of road infrastructure presents significant challenges to public safety and maintenance budgets, with cracks serving as critical indicators of structural instability. Despite extensive advancements, exist...