No abstract

No abstract

This paper describes a new versatile algorithm for correcting nonlinear distortions, such as curvature of book pages, in camera based document processing. We introduce the idea of using local orientation features to interpolate a vector field from which a warping mesh is derived. Ultimately, the image is corrected by approximating the nonlinear distortion with multiple linear projections. Since the algorithm does not derive the mesh directly from text baselines it is robust over arbitrarily complex text layouts. We describe a baseline detector for extracting the required local orientation features. We also sketch a method for correcting nonlinear distortions of a document's vertical axis with our algorithm.

Images taken under water are often of a monochromatic appearance, due to the physical interaction (absorption and reflection) between particles and light sources. Enhanced images with improved saturation, for which the monochromatic character has been corrected, are more suitable for generating 3D models and for identifying structures and materials by human experts. In this paper, we present an automatic method to identify the mean water color from a set of images. This mean color represents an average gray and is used to describe a new axis in CIELab color space. An extended color variance and a histogram equalization are simultaneously applied to the image. The main advantage of this method is the fully automatic enhancement process. An UUV (Unmanned Underwater Vehicle) can operate without providing a color reference scheme. The presented method was implemented in the software JEnhancer, which is freely available. JEnhancer was successfully tested in several documentation campaigns, and was integrated into the videogrammetric software pipeline Archaeo3D to produce 3D models from videos.

This paper presents a new algorithm for fusioning images of text-documents taken with different exposures. It is compared to several standard block oriented exposure-and focus-blending-algorithms. The recognition rate of a publicly available OCR-engine is used as a benchmark to quantify the results. Experiments show in average an improvement in the recognition rate from 0.46 to 0.64 by employing exposure blending as preprocessing step to an OCR. The presented algorithm of blending high-pass filtered images instead of original images further increases the recognition rate to 0.95.