If you can get by with decimation, the answer is simple.
Anthony de Vries on 8/16/2001 answered this question.
He basically unflattened the flattened image out of the "Read JPEG" and then removed rows and columns from the pixmap matrix to shrink the image. If your JPEG must be arbitrarily scaled (i.e. can't be a 1:2, 1:3, 1:4, etc. decimation), you will have to do as IreneHe suggested and interpolate the values to match "fractional indices" of the new array. You should be able to use the array functions to do this.
After you decimate your pixmap matrix, send it through the flatten function and draw the flattened pixmap to your picture indicator.
I'm sorry I don't have exact function names. I am away from my LabVIEW machine at the moment.
jc
Mac 10.4
LV7.1
CLD