VCT Vision ELITE Series Especificaciones Pagina 59

  • Descarga
  • Añadir a mis manuales
  • Imprimir
  • Pagina
    / 157
  • Tabla de contenidos
  • MARCADORES
  • Valorado. / 5. Basado en revisión del cliente
Vista de pagina 58
GE H
EALTHCARE
D
IRECTION
DOC0500644, R
EV
4 D
ISCOVERY
, L
IGHT
S
PEED AND
B
RIGHT
S
PEED
DICOM C
ONFORMANCE
S
TATEMENT
59 of 157
GEMS "PRI-CTMR-MOD12" DICOM MOD "JPEG" Encoding Information
Note: Some earlier versions of that were intended to write JPEG Lossless Process
Selection Value 14 1.2.840.10008.1.2.4.70 actually contained errors in byte
ordering of encapsulation of fragment item tags, selection of entropy coding
tables, and calculation of difference values from predictors. These errors cause
the compressed images to be unreadable by non-GEHC software. All future GEHC
LightSpeed software will read (but not write) archive media written with these
errors.
The early release product software had a JPEG encode routine that is called the
BAD_ENCODER for the following explanation.
In the BAD_ENCODER produced 'SOS' (0xFFDA) header, the entropy coding table
selector codes are 0x11. They should be 0x00. This mistake is the key that specifies the
BAD_ENCODER processing problems
The BAD_ENCODER computation of the pixel value difference to be entropy encoded is
computed as (Predictor - value) when it should be calculated as (value - Predictor)1. The
result is that the decompression via industry standard packages result in a negative of
the original image pixel value.
The BAD_ENCODER predictor value used at the beginning of each line uses the last
value of the previous line in the image. The JPEG standard says the predictor for the first
line element of each line should be the first element of the line immediately above the
current line. The first line, first element predictor should be the unsigned value that is half
the full-scale range for the "sample precision". The first line, first element predictor is
correctly determined in the BAD_ENCODER.
Explanation
This section is provided to explain what was done during Image encoding so that people who
wish and need to decode the images can modify the available JPEG codec software. These
packages are available from the WEB and were originally developed at Stanford and Cornell as
well as others in the industry.
What was done in Generation 1 DICOM MOD encoding by GEHC:
In the JPEG codec output, the JPEG 'SOS' header defines the Huffman table selector codes to
find the appropriate Huffman table (FF DA 00 08 01 00 11 01 00 00 ...). The GEHC DICOM
Access Kit coded these as 0x11. They should have been 0x00. Zero was the number assigned
in the "DHT" header. This error is a primary key that is used in the GEHC Generation 2 system
to determine that the JPEG data was encoded using the GEHC Generation 1 coding technique.
1 No specific direct confirmation of the (value - Predictor) policy was found, but the graphic on page 185 of the book "JPEG
still image data compression standard" by Pennebaker shows the predictor entering the adder as a negative element and
the value entering as positive element. Like wise, both Cornell and Stanford packages perform the computation as (value -
Predictor).
Vista de pagina 58
1 2 ... 54 55 56 57 58 59 60 61 62 63 64 ... 156 157

Comentarios a estos manuales

Sin comentarios