Gamma LAT: Reference Manual

float2short, float2uchar, short2float, uchar2float

ANSI-C programs: float2short.c, float2uchar.c, short2float.c, uchar2float.c

NAME
float2short - Convert file of type float to file of type short integer.
float2uchar - Convert file of type float to file of type unsigned char.
short2float - Convert file of type short integer to file of type float.
uchar2float - Convert file of type unsigned char to file of type float.

SYNOPSIS
float2short <infile> <outfile> [a] [b]

<infile>	input data file (float)
<outfile>	output data file (short integer)
[a]	scaling factor (default=1.0)
[b]	scaling exponent (default=1.0)

float2uchar <infile> <outfile> [a] [b]

<infile>	input data file (float)
<outfile>	output data file (unsigned char)
[a]	scaling factor (default=1.0)
[b]	scaling exponent (default=1.0)

short2float <infile> <outfile> [a] [b]

<infile>	input data file (short integer)
<outfile>	output data file (float)
[a]	scaling factor (default=1.0)
[b]	scaling exponent (default=1.0)

uchar2float <infile> <outfile> [scale] [exp [offset]

<infile>	input data file (unsigned char)
<outfile>	output data file (float)
[scale]	scaling factor (default=1.0)
[exp]	exponent (default=1.0)
[offset]	offset (default=0.0)

EXAMPLES
short2float short_integer_dem float_dem 0.1 1.0
(converts the input file of type short integer to floating point and multiplies each value with the scale factor 0.1)

float2short 8056.mli 8056.mli_short 1e03 0.5
(converts the input multi-look intensity image (float) to short integer format. For the scaling of the values the function 1e03*(input value)^0.5 is used, i.e. the square root of the input values is multiplied with 1000).short2float 8056.mli_short 8056.mli 1e-06 2.0
(converts the input multi-look intensity image (short integer) to float format. For the scaling of the values the function 1e-06*(input value)^2.0 is used, i.e. the square of the input values is multiplied with 1e-06).float2char 8056_8560.cc 8056_8560.cc_char 255. 1.0
(converts the input coherence image (float) to unsigned char format. For the scaling of the values the input values are multiplied with 255.0).

DESCRIPTION
float2short, float2uchar, short2float, uchar2float convert image data between 4-byte float, short integer (2-byte), and unsigned char (byte) formats. the uchar2float program also has the option to subtract a floating point number from the data prior to scaling, i.e. the ouput values are scale * (data_in - offset)**exp

The programs read the input data, apply the user defined scaling function of the type out = a*in^b and write out the scaled value in the output format.

Notice the valid ranges of the short integer and unsigned char formats:

format	value range
short integer	-32768 ... 32767
unsigned char	0 ... 255

Values which are outside the valid range after application of the scaling function are set to the nearest valid value.

The programs float2short and float2uchar allow to reduce the volume of image data by factors of 2 and 4, respectively.

Compression of SAR backscatter coefficients (files such as *.mli, *.pwr)
To reduce the data volume of SAR backscattering coefficients to 2 byte per value we recommend to use float2short with a=1e03 and b=0.5. short2float with a=1e-06 and b=2.0 allows to convert such compressed short integer format data back to float format.

To reduce the data volume of SAR backscattering coefficients to 1 byte per value we recommend to use float2uchar with a=177.8 and b=0.25. uchar2float with a=1e-09 and b=4.0 allows to convert back such compressed unsigned char format data back to float format.

Compression of coherence data (files such as *.cc, *.cc_ad)
To reduce the data volume of coherence data to 1 byte per value we recommend to use float2uchar with a=255.0 and b=1.0. uchar2float with a=0.00392 and b=1.0 allows to convert such compressed unsigned char format data back to float format.

OPTIONS
ESA processed SLC and PRI data can be converted from short integer to float format using short2float.

Decompression of ESA processed SLC data
To decompress ESA processed SLC data short2float with b=1.0 may be used. In addition the scale factor a can be used to achieve an approximate radiometric calibration of the data.

Decompression of ESA processed PRI data
To decompress ESA processed PRI data short2float with b=2.0 may be used (Notice that ESA PRI values are proportional to the square root of the backscatter intensity. The values of the MSP and ISP *.mli and *.pwr files, on the other hand, are proportional to the backscatter intensity).

SEE ALSO
LAT_file_formats