Discrete techniques for computer transformations of digital images and patterns

作者:

Highlights:

摘要

Images and patterns through a cycle conversion T−1T are discussed and facilitated by the combined algorithms, where the transformation is T:(ξ,η)→(x,y) with the linear or nonlinear functions x = x(ξ,η) and y = y(ξ,η). A new Area Method is presented for images through T and T−1T of linear transformations, and three combinations of the Splitting-Shooting Method and the Splitting-Integrating Method are proposed for images through T−1T of linear and nonlinear transformations. Furthermore, both error analysis and graphical experiments given prove the importance of those combinations to computer vision, image processing, graphs and pattern recognition.

论文关键词:Shape transformation,Nonlinear model,The splitting-shooting method,The splitting-integrating method,Combined algorithms,Multi-grey-image,Numerical methods,Error analysis,Computer algorithm,Computer vision,Image processing,Pattern recognition

论文评审过程:Received 29 September 1989, Revised 18 January 1990, Accepted 19 February 1990, Available online 19 May 2003.

论文官网地址:https://doi.org/10.1016/0031-3203(90)90120-A