The Interpolation-Based Bintree and its Application to Graphics

A new structure for representing binary images, called the Interpolation-Based Bintree is introduced. This structure combines the features of some existing representations such as linear quadtrees, binary trees, and interpolation-based codes, to improve the performance of operations manipulating graphics images. The implementation of this method is performed on both randomly generated and actual images and is shown to be very simple and efficient. It is also shown to compare favorably with several popular techniques in both storage and processing requirement. Although this structure is mainly for binary images, it can also be utilized for a class of “color-coded” images, where the color-code has a nominal value, the difference between two codes has no meaning, and where areas of uniform code value are larger with respect to resolution.