Seam Carving Algorithm for Image Adaptation Based on Distortion Control

To reduce the information loss and geometric distortion of image retargeting, a novel seam carving method for image retargeting based on image distortion control was proposed. In the proposed method, a saliency map combining local and global contrast saliency was used as the energy map for sorting out the seams need to be carved based on attention saliency, which can preserve the region of interest (ROI) better. Meanwhile, the dense SIFT flow was utilized to measure the degree of image distortion during seam carving. During seam carving, each time a certain number of seams were removed, and the distortion degree between the original image and the resized image was calculated. Once the distortion reached a certain threshold, seam carving was terminated and switched to scaling the image to the required size. The experimental results demonstrate that the proposed method keep a good balance between seam carving and scaling for image retargeting and is better at preserving the main content information and avoiding the distortion of key objects.