<p>This paper proposes an effective image and video enhancement framework that combines pixel scaling, guided filtering, and a Vision Transformer (ViT) to improve visual quality while preserving structural details. The approach begins with pixel scaling to suppress noise and normalize intensity variations, followed by guided filtering to enhance edges and fine structures. To further exploit spatial and temporal correlations, a ViT-based module is employed to capture long-range dependencies and global contextual information across frames. The proposed method is evaluated on the VideoSR dataset, which includes diverse real-world scenes with complex motion, occlusions, and varying texture characteristics. Experimental results demonstrate consistent improvements in both objective and perceptual quality metrics. The integration of pixel scaling and guided filtering yields PSNR gains of up to 0.44 dB across different images, indicating effective noise reduction and detail preservation. The ViT-based enhancement stage further boosts performance, achieving <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\Delta \)</EquationSource> <EquationSource Format="MATHML"><math> <mi mathvariant="normal">Δ</mi> </math></EquationSource> </InlineEquation> PSNR improvements of up to 1.83 dB and SSIM gains of up to 0.088, depending on the patch size and batch configuration. Among the evaluated settings, a patch size of 8<InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(\times \)</EquationSource> <EquationSource Format="MATHML"><math> <mo>×</mo> </math></EquationSource> </InlineEquation>8 offers an optimal balance between enhancement performance and computational efficiency. Overall, the results confirm that the proposed framework delivers robust and high-quality enhancement, making it suitable for practical image and video processing applications.</p>