A fast spatial patch blending algorithm for artefact reduction in pattern-based image inpainting

Abstract

Figure 1: Illustration of our proposed spatial patch blending algorithm for image inpainting. From left to right : color image with area to reconstruct, reconstruction result with the inpainting algorithm from [Criminisi et al. 2004], our reconstruction result. AbstractWe propose a fast and generic spatial patch blending technique that can be embedded within any kind of pattern-based inpainting algorithm. This extends our previous work on the visual enhancement of inpainting results. We optimize this blending algorithm so that the processing time is roughly divided by a factor ten, without any loss of perceived quality. Moreover, we provide an open-source and simple-to-use software to make this easily reproducible.

Tensor-Directed Spatial Patch Blending for Pattern-Based Inpainting Methods

“…In Fig. 8 we show comparisons of our method results (right column) with those of [12] (middle column). In a common way, one can see that the results of [12] present some structure superposition effect due to the blending of structures and flat areas, like between the rocks (rows 2-4).…”

“…The elemental formulation of the spatial patch blending, as pointed out in [12], is slow to compute and does not allow to use it easily. In this section, we propose an adaptation of the work of [12] to the geometric model proposed in this paper.…”

“…In this section, we propose an adaptation of the work of [12] to the geometric model proposed in this paper. The proposed algorithm remains quite the same that those in [12]. The first difference lies in the Gaussian weights used to apply patch blending at each scale of the blending amplitude map.…”

“…The first difference lies in the Gaussian weights used to apply patch blending at each scale of the blending amplitude map. While these weights are isotropic in [12], in this version we use anisotropic Gaussian weights based on the geometric model described in 2.1 Fig. 5.…”

“…Spatial Patch Blending: In our previous works [13,12], we proposed a patch blending method that tries to reduce block effect artifact in pattern-based image inpainting results. The core of the method is to mix the data of several patches in a way that the reconstruction seems more continuous than with blocky patch chunks.…”