Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

JPEG Print Object Switching, Using a JPEG Marker, in Object Oriented Computer Printing Systems

IP.com Disclosure Number: IPCOM000010643D
Original Publication Date: 2003-Jan-02
Included in the Prior Art Database: 2003-Jan-02
Document File: 2 page(s) / 50K

Publishing Venue

IBM

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 44% of the total text.

Page 1 of 2

  JPEG Print Object Switching, Using a JPEG Marker, in Object Oriented Computer Printing Systems

   In computer printing systems, it is normal for the hardware controlling the printer, or marking engine, to only hold in memory and operate on portions of a page in order to reduce the memory cost and reduce the required physical space for memory . The portions of data that may be handled and held at one time may be as small as one scan line (one laser beam path across the page on a laser printer) or may be multiple scan lines.

Furthermore, there are two types of printing system controllers. One is an object oriented controller which deals with objects and the placement of these objects on a page. Objects that may be much smaller than the page width or cover the whole page. The other type of printer controller is a raster oriented controller which receives data a scan line at a time. In raster oriented controllers the data arrives in order from the first PEL in a scan line to last PEL in a scan line. This is unlike the object oriented controller which receives object related data that usually is data for the left most object followed by data for the next left most object and so on. The placement of each object may be a distance from the previous object. Thus, the intervening data between the two objects is not transmitted and thus not received by the printer controller. The input data stream therefore, has data next to each other that is not necessarily next to each other on the page. Given further that the data for each object is in a JPEG compressed format it is impossible to tell where one scan line ends and where the next scan line starts because of the JPEG variable length compressed format.

Being unable to tell where a scan or multiple scan segment ends in the controller input data is a problem for object oriented controllers that work on only a portion of a page and therefore on only a portion of an object on the page. The reason this is a problem is because a controller working on a fixed number of scan lines, called a strip, is not allowed to completely read an object and processes it. This is because there is not enough memory to do so. The requirements for such a controller therefore must be to read in a few scan lines of an object, stop and process that data, then skip or switch to the next object and process the fixed number of scan lines in that object. It must stop the read in of data at the end of each strip (a fixed number of scan lines is defined as a strip) and switch to the read in of data for the next object, find that objects end of strip, stop, switch to next object, and so on.

Controllers without an end of strip detection scheme have the disadvantage that they can only do one of two things. It can read in one JPEG compressed word at a time, count the decompressed words that are generated, determine if the edge of object has been reach or passed. If the edge of the object is not detected it must get another word and...