-
公开(公告)号：DE3481966D1
公开(公告)日：1990-05-17
申请号：DE3481966
申请日：1984-10-31
Applicant: IBM
Inventor： ITOH MASAHARU , YANAGISAWA HIROSHI , YOKEMURA TAKETOSHI
-
公开(公告)号：CA1240789A
公开(公告)日：1988-08-16
申请号：CA467681
申请日：1984-11-13
Applicant: IBM
Inventor： ITOH MASAHARU , YOKEMURA TAKETOSHI , YANAGISAWA HIROSHI
Abstract: An image processing system is characterized in that a plurality of image processing devices having different resolutions are included. An original image to be sent from a selected image processing device is constituted by a base image corresponding to the lowest resolution, and one or more error compensation data corresponding to resolutions changing stepwise from the resolution of the base image to that of the original image. The base image is generated by reducing stepwise the resolution of the original image. The error compensation data is generated by predicting from either a lower resolution image whose resolution is lower than that of the original image, or, the base image, a higher resolution image whose resolution is one step higher, and by encoding error between the predicted image and an actual higher resolution image. When the original image is to be sent to another selected image processing device, image data only up to the one having a resolution of another device are sent.
-
公开(公告)号：DE69513071D1
公开(公告)日：1999-12-09
申请号：DE69513071
申请日：1995-03-16
Applicant: IBM
Inventor： GARFUNKEL GLEN ADAM , SALO MIKE PAUL , AOYAGI AKIHIKO , YANAGISAWA HIROSHI , TERASHIMA HIROSHI , KUROKI KENJI
Abstract: In a disk drive actuator head stack assembly, magnetic transducers including magnetoresistive (MR) read heads are connected to a current source and configured to provide an MR sensor current to each MR head wherein the direction of current flow for the MR sensor current in each MR head is the same relative to the actuator for all of the MR heads in the actuator head stack. This configuration allows all of the MR heads in the head stack assembly to initialized simultaneously by applying a homogeneous magnetic field to the head stack assembly, thus providing the capability of initializing the MR heads at the actuator or disk drive level rather than requiring that the MR heads be initialized individually.
-
公开(公告)号：DE3850029D1
公开(公告)日：1994-07-14
申请号：DE3850029
申请日：1988-09-16
Applicant: IBM
Inventor： ITOH MASAHARU , TOYOKAWA KAZUHARU , YANAGISAWA HIROSHI
Abstract: An original image of a text document is scanned by an image scanner of a resolution of 240 pel/inch in a terminal device. The image scanner generates binary data for each pel of the image sampled at the 240 pel resolution which are stored in a input image buffer. The pel signals are selectively fetched to generate a first compressed data stream 120 pel resolution and a second compressed data stream of the remaining pels. To selectively fetch the pel signals four sampling windows are prepared. The first window fetches one pel signal, as a current pel signal being processed, stored at one of every other pel position and a plurality of neighbouring pel signals stored at the every other pel positions. The sampling points of the first window are arranged at 120 pel resolution. Second, third and fourth sampling windows are used to fetch or sample both pel signals at 120 resolution and the remaining pel signals in various combinations. Patterns of the second, third and fourth windows differ from each other and are arranged at 240 pel resolution.
-
公开(公告)号：DE3687608D1
公开(公告)日：1993-03-11
申请号：DE3687608
申请日：1986-06-20
Applicant: IBM
Inventor： SUZUKI NAOSHI , YANAGISAWA HIROSHI , YOKEMURA TAKETOSHI
Abstract: The apparatus calculates the intensity P of picture points (pels) in a transformed image (filtered, enlarged, reduced) from the equation … … where Ii are the intensities of each of m surrounding pels (reference window) in the original image and alpha i the associated weight factors.To allow flexible changes of the conversion, a loadable table (3) is provided which contains, for selected pels in the transformed image (e.g. along a scanning line), the associated weight factors of the pels in the reference window as well as the displacement by which the reference window has to be shifted to calculate the next transformed image pel.
-
Patent Agency Ranking
公开(公告)号：DE3686521T2
公开(公告)日：1993-03-11
申请号：DE3686521
申请日：1986-06-13
Applicant: IBM
Inventor： ASADA SHIGEKI , YANAGISAWA HIROSHI
IPC: G06F3/00 , H04N1/028 , H04N1/04 , H04N1/17 , H04N1/19 , H04N3/14 , H04N5/335 , H04N5/341 , H04N5/372
Abstract: The invention is devised to start application of clock pulses to a shift register in a sensor unit of an image scanner in synchronization with a read signal from a processor in a method for controlling the image scanner so as to enable to easily and directly fetch video data outputted from the shift register into the processor as its input data without using a DMA (Direct Memory Access) method or a program loop method that requires special measures in aspects of hardware and software.In addition, the invention is devised to provide a data disposing cycle in the operation mode of the shift register in addition to the data reading cycle, so that, even when interruption from others causes the operation of the processor fetching the video data from the shift register to exceed a predetermined scanning time before completion of the operation, the data disposing cycle is inserted right after completion of the operation to dispose high level data that has been sensed for a long period of time by a photosensor of the sensor unit, and normal level data or image information properly representing an image being scanned is obtained.Furthermore, the invention reduces significantly processing load of the processor with a simple hardware constitution by adding the operation of a run length counter etc. so as to enable high speed linear data compression.
-
公开(公告)号：DE3686521D1
公开(公告)日：1992-10-01
申请号：DE3686521
申请日：1986-06-13
Applicant: IBM
Inventor： ASADA SHIGEKI , YANAGISAWA HIROSHI
IPC: G06F3/00 , H04N1/028 , H04N1/04 , H04N1/17 , H04N1/19 , H04N3/14 , H04N5/335 , H04N5/341 , H04N5/372
Abstract: The invention is devised to start application of clock pulses to a shift register in a sensor unit of an image scanner in synchronization with a read signal from a processor in a method for controlling the image scanner so as to enable to easily and directly fetch video data outputted from the shift register into the processor as its input data without using a DMA (Direct Memory Access) method or a program loop method that requires special measures in aspects of hardware and software.In addition, the invention is devised to provide a data disposing cycle in the operation mode of the shift register in addition to the data reading cycle, so that, even when interruption from others causes the operation of the processor fetching the video data from the shift register to exceed a predetermined scanning time before completion of the operation, the data disposing cycle is inserted right after completion of the operation to dispose high level data that has been sensed for a long period of time by a photosensor of the sensor unit, and normal level data or image information properly representing an image being scanned is obtained.Furthermore, the invention reduces significantly processing load of the processor with a simple hardware constitution by adding the operation of a run length counter etc. so as to enable high speed linear data compression.
-
公开(公告)号：IN167388B
公开(公告)日：1990-10-20
申请号：IN491MA1986
申请日：1986-06-26
Applicant: IBM
Inventor： SUZUKI NAOSHI , YANAGISAWA HIROSHI , YOKEMURA DAKETOSHI
Abstract: The apparatus calculates the intensity P of picture points (pels) in a transformed image (filtered, enlarged, reduced) from the equation … … where Ii are the intensities of each of m surrounding pels (reference window) in the original image and alpha i the associated weight factors.To allow flexible changes of the conversion, a loadable table (3) is provided which contains, for selected pels in the transformed image (e.g. along a scanning line), the associated weight factors of the pels in the reference window as well as the displacement by which the reference window has to be shifted to calculate the next transformed image pel.
-
公开(公告)号：SG28235A1
公开(公告)日：1996-04-01
申请号：SG1995000165
申请日：1995-03-25
Applicant: IBM
Inventor： AOYAGI AKIHIKO , TERASHIMA HIROSHI , YANAGISAWA HIROSHI , KUROKI KENJI
IPC: G11B5/127
-
公开(公告)号：DE3850029T2
公开(公告)日：1994-12-01
申请号：DE3850029
申请日：1988-09-16
Applicant: IBM
Inventor： ITOH MASAHARU , TOYOKAWA KAZUHARU , YANAGISAWA HIROSHI
Abstract: An original image of a text document is scanned by an image scanner of a resolution of 240 pel/inch in a terminal device. The image scanner generates binary data for each pel of the image sampled at the 240 pel resolution which are stored in a input image buffer. The pel signals are selectively fetched to generate a first compressed data stream 120 pel resolution and a second compressed data stream of the remaining pels. To selectively fetch the pel signals four sampling windows are prepared. The first window fetches one pel signal, as a current pel signal being processed, stored at one of every other pel position and a plurality of neighbouring pel signals stored at the every other pel positions. The sampling points of the first window are arranged at 120 pel resolution. Second, third and fourth sampling windows are used to fetch or sample both pel signals at 120 resolution and the remaining pel signals in various combinations. Patterns of the second, third and fourth windows differ from each other and are arranged at 240 pel resolution.
-
-
-
-
-
-
-
-
-
Search Results
27
records
(took 0.018 seconds)
