Efficient Transformation of an N by N Array of Picture Elements
Original Publication Date: 1985-Mar-01
Included in the Prior Art Database: 2005-Feb-18
The desire for parallelism to achieve high performance in graphics and image processing displays has led to the design of frame buffers which provide parallel access to N by N square arrays of picture elements (pixels). The desire to avoid word boundaries during the parallel processing of the N by N square leads to a special memory system which allows the parallel access of any arbitrary square by providing different addresses to different memory chips. However, the bits read have to be circularly shifted in both the horizontal and vertical directions to allow them to be aligned with the processing array. A pixel transformer which provides the two-dimensional circular shifting, mirroring about both the x and y axes, as well as allows the transformation of the axes is described herein.