Abstract:
Edges and structures are the most essential parts of an image that contain several levels of information.
Achieving an edge preserving and texture smoothing filtered output that is close
to human vision is a significant part of image decomposition. Researchers working on edge
preserving filters have so far not focused on the area of finding scale aware local operations
practically. Current methods of image decomposition for edge preserving presume details
of image as a variant of low contrast, therefore such techniques apply filters that can extract
features as well as increase the contrast of images. To attain strong image smoothing and
edge preserving, an efficient edge preserving filtering technique is being proposed. An improved
edge preserving texture smoothing technique for high contrast images is developed.
The high contrast images separated based on histogram pattern are passed through a median
filter. The pre-processed image is divided into sub-images using quad-tree decomposition.
Edge preserving joint bilateral filter is then applied separately on each sub-image/block to
preserve the small edges. Simulation results compared with different state-of-the-art techniques
verify the significance of proposed technique.
Image aspect ratios are changed by content aware image retargeting techniques. These techniques
preserves visually prominent features of images. Seam carving is a simple image
operator that supports reduction and expansion by using content aware image resizing. For
image retargeting the proposed technique uses a smoothing operator based on covariance
matrix. The proposed approach captures the local texture and structure information by using
second order statistics as patch descriptors. The approach supports different visual saliency
measures for defining the energy of an image. The results Demonstrates that new improved
way enhances the resizing effects, avoids the distortion and produce better outcomes. Visual and quantitative comparison (with state of art existing technique) are performed to verify the signicance of proposed technique. Simulation results reveal that the proposed technique is almost two times efficient and more accurate as compared to state of art image edge preserving technique.
