yimage 1.0.1

A friendly remote sensing data library

Author: Xuan Yang

E-mail: yangxuan.ac#gmail.com

yimage is a friendly remote sensing data library for Python. It encapsulates OpenCV library and GDAL library, and uses OpenCV data format as a standard to provide a convenient remote sensing data I/O interface.

Installation

pip install yimage

Prerequisites

• Numpy

  • pip install numpy

• OpenCV

  • pip install opencv-python

• GDAL

  • pip install GDAL

  • conda install GDAL (We recommand this one)

  • (if missing libpoppler.so when import gdal, try running conda install "poppler<0.62")

• Cython

  • pip install Cython

Document

<package> yimage

use import yimage to import this package.

use yimage.__version__ to check the version.

<class> io

yimage.io.read_image(filename, driver='GDAL', read_range=None, with_geo_info=False)

Reading an image file from disk by GDAL or OpenCV.

image array format: (H, W) or (H, W, C). band order: (b, g, r, nir, …)

Return: [numpy array] or [numpy array, geo info dict]

filename: [filename]

driver: [GDAL], [OpenCV]

read_range: [None] if you want to load the whole image, [(xoff, yoff, xsize, ysize)] if you want to load the part of image (same parameters as ReadAsArray method of GDAL library)

with_geo_info: [False] only returns the numpy array, [True] returns numpy array and geo info dict (‘proj’ and ‘coord’)

img_ordered_by_band_list(img, band_list)

Reordering bands by a list.

Return: [reordered numpy array]

img: [numpy array]

band_list: [list] e.g. [1, 2, 3] for true color image, [2, 3, 4] for pseudo color image

write_image(filename, img, driver, dtype='uint8', gdal_driver='GTiff', img_geo_info=None, color_table=None, no_data_value=None)

Writing an image file to disk by GDAL or OpenCV.

image array format: (H, W) or (H, W, C). band order: (b, g, r, nir, …)

Return: [True] for success, [False] for failed.

filename: [filename]

img: [numpy array]

driver: [GDAL], [OpenCV]

dtype: [uint8], [uint16], [int16], [uint32], [int32], [float32], [float64]

gdal_driver: [Gtiff] GeoTIFF format, [ENVI] ENVI Standard foemat, [HFA] ERDAS format, and also, supports all formats that supported by GDAL library, please see the docs of GDAL for more driver name string.

img_geo_info: [None], [geo info dict] dict (‘proj’ and ‘coord’)

color_table: [None], [color tabel list] for coloring single band image

no_data_value: [None], [an integer] for setting NoDataValue

polygonize(raster_filename, shapefile_filename, band=None, layer_name=None)

Polygonize a raster data.

Return: [True] for success, [False] for failed.

raster_filename: [filename for raster data]

shapefile_filename: [filename for ESRI shapefile]

band: [None] polygonize the 1st band, [an integer] polygonize that band

layer_name: [None] the layer name of shapefile will be ‘polygon’, [name string]

load_color_table_file(filename)

Load color table from file.

Return: [color table list]

filename: [filename for color table file]

file format: one color RGB value with comment per line. Demo:

255 / 0 / 0 # Color 1

0 / 255 / 0 # Color 2

0 / 0 / 255 # Color 3

Updates

Version 1.0.1 (06/12/2020)

• offset method in write_image is deprecated

Version 1.0.0 (06/12/2020)

• Support reading and writing remote sensing data

• Support reordering bands of data

• Support polygonizing the raster data

• Support processing color table file