Wrong subset Y with mages without georeferenced

[Bang Pham Huu]

I could ingest georeferenced TIFF file with latitude and longtitude but could not ingest iungeoreferenced PNG like rgb, mr,… because

http://localhost:8080/rasdaman/ows?service=WCS&version=2.0.1&request=UpdateCoverage&coverageId=ad09&subset1=i(0.0,400.0)&inputCoverageRef=file:///tmp/90008feb_64a4_43b9_b486_bd92a90be50b.gml&subset2=j(344.0,0.0)

in the subset2 it shows a wrong values, it should be miny, maxy: j(0, 344). I've investigated in wcst_import/util/gdal_util.py and see your formula to calculate min, max of X and min, max of Y.

See your code first

# My comment for better understanding
 GDAL Geotransform:
          + 0: top left x
          + 1: west-east pixel resolution (x axis)
          + 2: 0 value
          + 3: top left y
          + 4: 0 value
          + 4: north-south pixel resolution (y axis)

        geo_transform = self.gdal_dataset.GetGeoTransform()
        minx = geo_transform[0]
        maxy = geo_transform[3]
        maxx = minx + geo_transform[1] * self.gdal_dataset.RasterXSize
        miny = maxy + geo_transform[5] * self.gdal_dataset.RasterYSize

It is right with the case when the anchor of axis is in the center (0, 0) because with minx=geo_transform[0] (top left X), maxy = geo_transform[3] (top left Y) and from this point you can canculate maxx and minyy below (low right).

*minx, maxy

*maxx, miny

←————-(0,0)————————→

But with ungeoreferenced image (the anchor point is in the top left) then minx = maxy = 0.0 (it is not TRUE).

• minx, *maxy (0,0)————→ x

|
|
|
|
V y

so that is why I has problem with import RGB images to Petascope because "j" subset is wrong order (max first and min after, like 344.0, 0.0).

I could fix this problem.

[Dimitar Misev]

Great that you found the bug Bang, best if you just fix it right away.

[Bang Pham Huu]

yes, I've fixed it Dimitar, thanks, I will submit patch now

 # BangPH check what kind of images is imported (Geo-Referenced or UnGeo-Referenced first)
        geo_transform = self.gdal_dataset.GetGeoTransform()
        if geo_transform[0] == geo_transform[2] == 0: # UnGeo-Referenced image
            minx = 0
            miny = 0
            maxx = geo_transform[1] * self.gdal_dataset.RasterXSize # Calculate the maxx = resolution * MaxWidth
            maxy = geo_transform[5] * self.gdal_dataset.RasterYSize # Calculate the maxy= resolution * MaxHeight
        else: # Geo-Referenced image
            minx = geo_transform[0]
            maxy = geo_transform[3]
            maxx = minx + geo_transform[1] * self.gdal_dataset.RasterXSize
            miny = maxy + geo_transform[5] * self.gdal_dataset.RasterYSize
        return {
            "x": (minx, maxx),
            "y": (miny, maxy)
        }

[Alex Dumitru]

This is a duplicate of ticket #1021