Structural VariationBased Adaptive LowLight Video Enhancement with 3D Median Filtering

This paper presents an advanced video frame enhancement and extension system that leverages adaptive im- age processing techniques to improve video quality in low- light and sand-dust conditions. The system processes video frames through a multi-step pipeline involving structural variation-based (SV) frame selection and adaptive bright- ness adjustment. Frames requiring enhancement undergo Adaptive Dual-Band Compensation (ADBC) for brightness correction and detail sharpening through Rolling Guided Filtering (RGF) and Guided Image Filtering (GIF). An extension technique further improves frame quality by applying histogram stretching, sharpening using the imsharpen function, and noise removal using 3D median filtering (medfilt3). The proposed method improves video clarity, contrast, and sharpness while reducing noise, resulting in a more refined and visually enhanced output suitable for surveillance and multimedia applications. Key Terms – Low-light sand-dust video, adaptive dynamic brightness correction, rolling guidance filter, dual-threshold interframe detection strategy, 3D median filtering, structural variation, guided image filter.