Local Water-Filling Algorithm for Shadow Detection and Removal of Document Images

Abstract

Shadow detection and removal is an important task for digitized document applications. It is hard for many methods to distinguish shadow from printed text due to the high darkness similarity. In this paper, we propose a local water-filling method to remove shadows by mapping a document image into a structure of topographic surface. Firstly, we design a local water-filling approach including a flooding and effusing process to estimate the shading map, which can be used to detect umbra and penumbra. Then, the umbra is enhanced using Retinex Theory. For penumbra, we propose a binarized water-filling strategy to correct illumination distortions. Moreover, we build up a dataset called optical shadow removal (OSR dataset), which includes hundreds of shadow images. Experiments performed on OSR dataset show that our method achieves an average ErrorRatio of 0.685 with a computation time of 0.265 s to process an image size of 960×544 pixels on a desktop. The proposed method can remove the shading artifacts and outperform some state-of-the-art methods, especially for the removal of shadow boundaries.

Document Shadow Removal with Foreground Detection Learning From Fully Synthetic Images

“…Existing works on shadow removal from a single document have proposed real datasets that contain pairs of shadow images and shadow-free images. Nevertheless, due to the cost of creating large datasets, there are only small real datasets available [1,3,5,7]. To combat this problem, [7] created a relatively large dataset by synthesizing shadows to the document using a graphic renderer.…”

“…Although several real image datasets have been proposed for document shadow removal [1,3,5,7], the total number of pairs is less 1,000. Even the largest publicly available real image dataset contains less than 100 unique scenes.…”

“…Most existing traditional document shadow removal methods use heuristics for exploring characteristics of document images [1,2,3,4,5]. In [1], a technique was proposed that detects intensity change in images and enforces a consistent background color to produce an unshadowed output.…”

Image Segmentation in Shape Synthesis, Shape Optimization, And Reverse Engineering

“…Segmentation of geometry is one of the basic requirements for reverse engineering, shape optimization, and shape synthesis. There are successful algorithms for segmenting 3D geometry (3D point cloud / triangulated surface) [1,2,3], and successful algorithms for segmenting 2D images [4,5,6,7,8]. Algorithms for both dimensions, 2D and 3D, are based on the same ideas/approaches.…”

“…The figure below (Figure. 2) presents the surface of the parametric projection domain (Figure 1 b).In this image, we apply the water shadow algorithm[8] after the iterative distance algorithm.…”