Adaptive Image Interpolation using a Multilayer Neural Network
Common image interpolation methods apply an interpolation function indiscriminiately to the whole image. The resulting image in general suffers from edge blurring and aliasing, which is a particular problem for the text regions of electronic documents.
In this paper we present a resizing neural network for edge and detail preserving image interpolation. The multilayer neural network is trained by using pairs of high resolution and low resolution imagery.
The high resolution is an 8-bit image scanned at 600 dpi. The low resolution image (300 dpi) is either a processed version of the high resolution image, or it is scanned independently. Local texture and edge features are extracted from the low (high) resolution image and are used as inputs to the neural networks. The interpolated pixels obtained as output are compared with the high (low) resolution pixels and the error is used to train the neural network.
Simulation results show that the proposed approach produces noticebaly sharper edges than traditional interpolation techniques including bilinear and bicubic interpolation.
Document Type: Research Article
Publication date: 2000-01-01
For more than 30 years, IS&T's series of digital printing conferences have been the leading forum for discussion of advances and new directions in 2D and 3D printing technologies. A comprehensive, industry-wide conference that brings together industry and academia, this meeting includes all aspects of the hardware, materials, software, images, and applications associated with digital printing systems?particularly those involved with additive manufacturing and fabrication?including bio-printing, printed electronics, page-wide, drop-on-demand, desktop and continuous ink jet, toner-based systems, and production digital printing, as well as the engineering capability, optimization, and science involved in these fields. In 2016, the conference changed its name formally to Printing for Fabrication to better reflect the content of the meeting and the evolving technology of printing.
Please note: For purposes of its Digital Library content, IS&T defines Open Access as papers that will be downloadable in their entirety for free in perpetuity. Copyright restrictions on papers vary; see individual paper for details.
- Information for Authors
- Submit a Paper
- Subscribe to this Title
- Membership Information
- Terms & Conditions
- Ingenta Connect is not responsible for the content or availability of external websites