A combination of normalized cross-correlation and least-squares matching is used. Since the latter needs good starting values, an image pyramid strategy is used starting at the level of lowest resolution with a one dimensional cross correlation process that works without operator assistance. Correlation results are considered approximate values in the next pyramid level which makes matching a totally automatic process. On each level the point raster is densified until the bottom level is reached. In order to keep computation time short, only in this original image the least-squares method is used to provide sub-pixel results. The found matches und ergo a quality control, where several criteria are tested and it is decided which results are accepted. Recently, the two steps that consume the Hon's share of computation time, Le. resampling and matching, are sped up considerably by the mutation of the DSS into an Advanced DSS. In the actual constellation this system uses a 14-transputer Paracom MultiCluster hosted on a SUN workstation.

http://www.isprs.org/proceedings/XXIX/congress/part5/610_XXIX-part5.pdf

Your question is not quite clear. Would you like to have a simple statistics on the size of the drops in the image or would would you like to know the distribution of the drops geografically in the image?

In any case it is helpful to apply an morphological filter which detects drops of different size. This is a standard operation in Matlab where you only have to define the structure element SE.Or you apply a  so called Hit-or-Miss-Filter which detects the drops in disfferent sizes. You get an amount of images with all the drops of diffrent sizes whicxh you can use for a histogram analysis. You can then calculate some useful statistics globally or locally.