Fixed issue # 17 (Coregistration sometimes fails in case of floating point coordinates of the input images.)

Fixed issue #23 ('TypeError in case COREG or COREG_LOCAL is called with an in-memory reference or target image and path_out is set to 'auto'.').

CI setup now updates ci_env environment installed via docker_pyenvs instead of creating an independent environment.

CI Python environment is not separate from the base env. Fixed mixed channels for gdal and libgdal causing libkea issue during CI.

uses majority filter to smooth output

try get peak of SCPS in aspect of dem
possilbe improvement for large landslide areas (test iran EQ landslide with alos 30m dem)

Conflicts:
	tests/test_COREG.py