This paper studies an image inpainting solution based on a primal sketch representation model [1], which divides an image into structure (sketchable) and texture (non-sketchable) components. This solution first predicts the missing structures, such as curves and corners, using a tensor voting algorithm [2]. Then the texture parts along structural sketches are synthesized with the patches sample...

It has been widely believed since Helmholtz that tangent discontinuities in image contours, such as T-junctions or L-junctions, will occur when one object occludes another. Here we describe a class of occlusion relationships where changes in 'border ownership' and amodal completion take place in the absence of tangent discontinuities in the image. We propose that even subtle curvature discontin...

The goal of our work is to complete the depth channel of an RGB-D image. Commodity-grade depth cameras often fail to sense depth for shiny, bright, transparent, and distant surfaces. To address this problem, we train a deep network that takes an RGB image as input and predicts dense surface normals and occlusion boundaries. Those predictions are then combined with raw depth observations provide...

Image inpainting is not only the art of restoring damaged images but also a powerful technique for image editing e.g. removing undesired objects, recomposing images, etc. Recently, it becomes an active research topic in image processing because of its challenging aspect and extensive use in various real-world applications. In this paper, we propose a novel e cient approach for high-quality and ...

Image completion has achieved significant progress due to advances in generative adversarial networks (GANs). Albeit natural-looking, the synthesized contents still lack details, especially for scenes with complex structures or images with large holes. This is because there exists a gap between low-level reconstruction loss and high-level adversarial loss. To address this issue, we introduce a ...

We present a new image completion method based on an additional large displacement view (LDV) of the same scene for faithfully repairing large missing regions on the target image in an automatic way. A coarse-to-fine distortion correction algorithm is proposed to minimize the perspective distortion in the corresponding parts for the common scene regions on the LDV image. First, under the assump...

Analyzing the limitations which exist in content-based image retrieval (CBIR) systems on optimization model, generalization design and retrieval efficiency, a combinative model based on image features correlation retrieval is designed, using some proper algorithms of image retrieval, relevance feedback principle and the advantage of DSP chip in image signal processing. The work mode is useful f...

User-provided textual tags of web images are widely utilized for facilitating image management and retrieval. Yet they are usually incomplete and insufficient to describe the whole semantic content of the corresponding images, resulting in performance degradations of various tag-dependent applications. In this paper, we propose a novel method denoted as DLSR for automatic image tag completion v...