How to get the pixels from a geotiff file in gdal python for a given polygon?

Question

I use python and GDAL under apache to make a webservice.

I have a GeoTIFF file and i want only a subset of the data, everything inside one polygon defined by a buffer of a line (coord is the input parameter of the service):

line = ogr.Geometry(ogr.wkbLineString)
for i in range(len(coord)):
    line.AddPoint(coord[i][0], coord[i][1])

line.Transform(coordTransform)    
polygon = ogr.Geometry(ogr.wkbPolygon)
polygon = line.Buffer(10.0, 2)

I find only solutions with shapefile, not with programmatically created geometries. Any help?

Answer 1

solved with:

[...]    
srcArray = gdalnumeric.LoadFile(rasterFile)

pxWidth = int(lrX - ulX)
pxHeight = int(lrY - ulY)

points = []
pixels = []
pts = poligono.GetGeometryRef(0)
for p in range(pts.GetPointCount()):
    points.append((pts.GetX(p), pts.GetY(p)))
for p in points:
    pixels.append(conf.world2Pixel(geoTrans, p[0], p[1]))

for i in range(len(pixels)):
    pixels[i] = (pixels[i][0]-ulX, pixels[i][1]-ulY)

rasterPoly = Image.new("L", (pxWidth, pxHeight), 0)
rasterize = ImageDraw.Draw(rasterPoly)
rasterize.polygon(pixels, 1)
mask = conf.imageToArray(rasterPoly)

valTiff = []
for r in range(pxHeight):
    for c in range(pxWidth):
        if mask[r][c]==1:
            valTiff.append(srcArray[r+ulY][c+ulX])