Abstract:
The guided filter and its derivatives are widely used in image processing and computer vision due to their low complexity and good edge-preservation. Despite this success, the guided filter variants cannot handle more aggressive filtering strengths, resulting in “detail halos”.
When input and guide images have structural inconsistencies, these filters perform poorly. In this paper, we show that these limitations are caused by the guided filter acting as a variable-strength locally-isotropic filter that weakly anisotropizes the image.
The adaptive guided filter (AGF), weighted guided image filter (WGIF), and gradient-domain guided image filter (GGIF) use unweighted averaging in their final steps, which causes this behavior.
We propose the Anisotropic Guided Filter (AnisGF), which uses weighted averaging to maximize diffusion while preserving strong image edges. To achieve strong anisotropic filtering with the low computational cost of the guided filter, the weights are optimized based on local neighbourhood variances.
Synthetic tests show that the proposed method addresses detail halos and inconsistent structures in guided filter variants. Scale-aware filtering, detail enhancement, texture removal, and chroma upsampling show the technique’s benefits.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
