Hazedefy: A Lightweight Real-Time Image and Video Dehazing Pipeline for Practical Deployment

TL;DR

Hazedefy addresses real-time dehazing on CPU-constrained devices by blending a physics-based Dark Channel Prior framework with lightweight, stability-focused approximations. It uses gamma-adaptive radiance reconstruction, a fast transmission estimate with lower bounds, and fractional top-pixel averaging for robust atmospheric light estimation, plus an optional guided-filter refinement for image-mode quality. The contributions include a practical, open, and reproducible pipeline that delivers real-time performance on consumer hardware without GPUs, suitable for mobile and embedded vision systems. Overall, Hazedefy offers a transparent baseline for application-driven dehazing that balances physical priors with engineering constraints for deployment in diverse real-world scenarios.

Abstract

This paper introduces Hazedefy, a lightweight and application-focused dehazing pipeline intended for real-time video and live camera feed enhancement. Hazedefy prioritizes computational simplicity and practical deployability on consumer-grade hardware, building upon the Dark Channel Prior (DCP) concept and the atmospheric scattering model. Key elements include gamma-adaptive reconstruction, a fast transmission approximation with lower bounds for numerical stability, a stabilized atmospheric light estimator based on fractional top-pixel averaging, and an optional color balance stage. The pipeline is suitable for mobile and embedded applications, as experimental demonstrations on real-world images and videos show improved visibility and contrast without requiring GPU acceleration.

Figures (2)

• Figure 1: Atmospheric scattering model. $I(x)$ is the observed hazy image, $J(x)$ is the scene radiance, $A$ is atmospheric light and $t(x)$ is the transmission.
• Figure 2: Flowchart of Hazedefy system.