Adding Depth to Cartoons Using Sparse Depth (In)equalities

Abstract

This paper presents a novel interactive approach for adding depth information into hand-drawn cartoon images and animations. In comparison to previous depth assignment techniques our solution requires minimal user effort and enables creation of consistent pop-ups in a matter of seconds. Inspired by perceptual studies we formulate a custom tailored optimization framework that tries to mimic the way that a human reconstructs depth information from a single image. Its key advantage is that it completely avoids inputs requiring knowledge of absolute depth and instead uses a set of sparse depth (in)equalities that are much easier to specify. Since these constraints lead to a solution based on quadratic programming that is time consuming to evaluate we propose a simple approximative algorithm yielding similar results with much lower computational overhead. We demonstrate its usefulness in the context of a cartoon animation production pipeline including applications such as enhancement, registration, composition, 3D modelling and stereoscopic display.

Micro perceptual human computation for visual tasks

“…Towards this end, we propose an HC algorithm that separates an image into layers using human computation. Our goal is similar to several previous works: we aim to be more robust than [Hoiem et al 2005;Assa and Wolf 2007;Saxena et al 2009], which use automatic methods, and, in contrast to [Oh et al 2001;Ventura et al 2009;Sýkora et al 2010], we do not require an expert user.…”

Repurpose 2D Character Animations for a VR Environment Using BDH Shape Interpolation

“…In the same year, Walt Disney Pictures, in the film Treasure Planet [14], presents a hybrid character, namely a 2D character with 3D components. Sỳkora and colleagues [29] introduce a method to allow users to specify depth inequalities in some sections of a 2D character, in order to generate a 2.5D popup. This technique has different objectives, such as enhancing the perception of depth in a 2D character, producing 3D-like shading or even stereoscopic images.…”

EXCOL: An EXtract-and-COmplete Layering Approach to Cartoon Animation Reusing

“…To make things worse, there possibly exists severe self-occlusion on the same layer, e.g., legs in Figure 6 (b-III), which makes completion of foreground layers rather challenging. Inspired by recent works on cartoon image registration [8] and depth reconstruction [34], we propose a global-to-local warping approach to sequentially warp a completed foreground layer L r k at frame k to help complete the corresponding layer L r k+1 in the next frame, in which auxiliary depth information can be optionally combined to guide the correct warping transformation.…”

“…Otherwise, artifacts like that shown in Figure 7 (d)) would appear. The depth can be added by using the technique proposed by Sýkora et al [34] in the key frame. But instead of setting (in)equalities over all the patches, we only need to specify inequalities on the self-occlusion regions.…”