@conference {Mansfield2012, title = {Visibility maps for improving seam carving}, booktitle = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)}, volume = {6554 LNCS}, number = {PART 2}, year = {2012}, pages = {131{\textendash}144}, abstract = {In this paper, we present a new, improved seam carving algorithm. Seam carving efficiently removes pixels from an image to produce a retargeted image. It has proved popular with users and has been used as a component in many retargeting algorithms. We introduce the visibility map, a new framework for pixel removing image editing methods. This allows us to cast retargeting as a binary graph labelling problem. We derive a general algorithm which uses seam carving operations for efficient greedy optimization of a well defined energy, and compare this with forward energy seam carving and shift map image editing. We test this method with varying parameters on a large number of images, and present an improved seam carving algorithm which can demonstrably produce better results. We draw general conclusions about pixel removing methods for retargeting and motivate future directions of research. {\textcopyright} 2012 Springer-Verlag.}, isbn = {9783642357398}, issn = {03029743}, doi = {10.1007/978-3-642-35740-4_11}, url = {http://www.adobe.com/products/photoshop/photoshopextended/features/}, author = {Mansfield, Alex and Gehler, Peter and Van Gool, Luc and Carsten Rother} } @conference {Mansfield2010, title = {Scene carving: Scene consistent image retargeting}, booktitle = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)}, volume = {6311 LNCS}, number = {PART 1}, year = {2010}, pages = {143{\textendash}156}, abstract = {Image retargeting algorithms often create visually disturbing distortion. We introduce the property of scene consistency, which is held by images which contain no object distortion and have the correct object depth ordering. We present two new image retargeting algorithms that preserve scene consistency. These algorithms make use of a user-provided relative depth map, which can be created easily using a simple GrabCut-style interface. Our algorithms generalize seam carving. We decompose the image retargeting procedure into (a) removing image content with minimal distortion and (b) re-arrangement of known objects within the scene to maximize their visibility. Our algorithms optimize objectives (a) and (b) jointly. However, they differ considerably in how they achieve this. We discuss this in detail and present examples illustrating the rationale of preserving scene consistency in retargeting. {\textcopyright} 2010 Springer-Verlag.}, isbn = {3642155480}, issn = {16113349}, doi = {10.1007/978-3-642-15549-9_11}, url = {www.fujifilm.com/products/3d/camera/finepix_}, author = {Mansfield, Alex and Gehler, Peter and Van Gool, Luc and Carsten Rother} }