Hi,

i want to calculate the ‘row projection’ vector for a given one-channel (8bit or float) image with dimensions ‘w’ and ‘h’.

The vector will be of size ‘h’, and the i-th entry of the vector is calculated as the sum of all pixels in the row ‘i’.

Is this somehow possible with the CUDPP library (or other free available libraries) ? I know there is reduction functionality in the CUDPP, but I’m not sure whether it is possible to do this ‘partial reduction’.

best regards and thx, Hannes