摘要:
As set forth herein, computer-implemented methods and systems facilitate watermarking documents and/or images using 2nd generation stochastic halftoning. The watermark is used to spatially vary the gray level at which a frequency modulation-to-amplitude modulation transition occurs. The encoding algorithm uses as inputs a contone image and a watermark. The visibility of the watermark is controlled by the magnitude of the difference between the AM-to-FM transition threshold values.
摘要:
As set forth herein, computer-implemented methods and systems facilitate halftoning by adapting 2nd generation halftone techniques to the spatial frequency content of an image utilizing a spot function introduced in conjunction with a technique for determining seed locations, which results in halftone images exhibiting improved edge and detail rendition. A dominant orientation of pixels in an input image is employed to select a pre-generated stochastic screen used to determine the seed locations.
摘要:
As set forth herein, computer-implemented methods and systems facilitate watermarking documents and/or images using 2nd generation stochastic halftoning. The watermark is used to spatially vary the gray level at which a frequency modulation-to-amplitude modulation transition occurs. The encoding algorithm uses as inputs a contone image and a watermark. The visibility of the watermark is controlled by the magnitude of the difference between the AM-to-FM transition threshold values.
摘要:
As set forth herein, computer-implemented methods and systems facilitate halftoning by adapting 2nd generation halftone techniques to the spatial frequency content of an image utilizing a spot function introduced in conjunction with a technique for determining seed locations, which results in halftone images exhibiting improved edge and detail rendition. A dominant orientation of pixels in an input image is employed to select a pre-generated stochastic screen used to determine the seed locations.
摘要:
As set forth herein, computer-implemented methods and systems facilitate halftoning using a parametrically controlled spot function based on triangle tessellation, which in turn facilitates dot growth control of periodic halftone using an irregular seed structure. The spot function determines the shape of the halftone dot used to reproduce a given pixel. The spot function is well suited for growing halftone dots arrayed on a non-regular grid and can also be used for grids with regular order. The spot function includes adjustable parameters for controlling its sharpness and slope.
摘要:
When restoring a scanned halftone image of a damaged document, parameters of the halftone structure (geometry) are estimated, as is an intended contone gray level for damaged portions of the image. Locations of halftone cell tiles in the scanned image are determined based on the halftone structure, and a halftone profile model is generated using the estimated halftone parameters and contone information. Image segments are aligned using global features for coarse alignment and halftone structure for fine alignment, such as in cases where the scanned halftone image contains unconnected fragments. Missing parts of the damaged document image are filled in using the halftone profile model and estimated contone values.
摘要:
As set forth herein, computer-implemented methods and systems facilitate halftoning using boundaries and centers of a polygonal tiling with a parameterized spot function that operates within the tiles. Defining the halftone structure includes defining the polygonal tiling with a specification of the polygon boundaries and center, and setting and applying parameters of the spot function, which utilizes center-to-boundary distances. The tiling can be defined explicitly, by defining a tile structure, or providing centers, one per polygon, and vertices for the polygons. The vertices and centers are used to generate the polygon boundaries. The polygonal tiling can be regular (e.g., repetitive) or irregular, and can also be varied in a manner adapted to the image content or to data that is being embedded.
摘要:
When restoring a scanned halftone image of a damaged document, parameters of the halftone structure (geometry) are estimated, as is an intended contone gray level for damaged portions of the image. Locations of halftone cell tiles in the scanned image are determined based on the halftone structure, and a halftone profile model is generated using the estimated halftone parameters and contone information. Image segments are aligned using global features for coarse alignment and halftone structure for fine alignment, such as in cases where the scanned halftone image contains unconnected fragments. Missing parts of the damaged document image are filled in using the halftone profile model and estimated contone values.
摘要:
As set forth herein, computer-implemented methods and systems facilitate employing a single stochastic screen defined up to a specified transition level. The dots at the specified level are connected to form a triangle tessellation with an associated spot function. For up to three colorants, a vector screening technique is employed, which involves a first colorant dot being grown from the vertices, second colorant dot grown from the triangle center, and a third colorant dot grown from the side midpoints. Growth in this manner ensures separation of colorants up to a certain density level. In cases where fourth colorant is to be used at a given pixel, the fourth colorant can be added using methods typically used for clustered screens, such as stochastic screening of the lightest colorant, in order to mitigate color shift sensitivity issues.
摘要:
As set forth herein, computer-implemented methods and systems facilitate halftoning using boundaries and centers of a polygonal tiling with a parameterized spot function that operates within the tiles. Defining the halftone structure includes defining the polygonal tiling with a specification of the polygon boundaries and center, and setting and applying parameters of the spot function, which utilizes center-to-boundary distances. The tiling can be defined explicitly, by defining a tile structure, or providing centers, one per polygon, and vertices for the polygons. The vertices and centers are used to generate the polygon boundaries. The polygonal tiling can be regular (e.g., repetitive) or irregular, and can also be varied in a manner adapted to the image content or to data that is being embedded.