Technical Abstract: It is of interest for many purposes, including nowcasting, to evaluate the structure of radar images in an effort to produce more accurate estimates of rainfall totals from radar data. Although subjective analysis can reliably determine the structure of radar imagery, computational techniques exist to analyze a radar image using algorithms that can be automated. Many of these techniques use some form of multiresolution analysis or Fourier analysis to accomplish structure identification.   One method of identifying structures of differing sizes and scales within an image is to use high, low, and band pass filters to highlight features of interest. This is frequently accomplished by decomposing an image into the frequency domain using a transform and performing the filtering operation within the frequency domain. A variety of transforms exist for accomplishing this including the Wavelet Transform (WT) and many transforms related to the Fourier Transform (FT).   An algorithm, based on the FT and Gaussian filters, has been developed to analyze a radar image and identify structures within the image. A variety of cases are then presented to demonstrate the performance and robustness of this algorithm. Primarily this algorithm delineates regions of convective and stratiform precipitation and identifies convective structures on multiple scales. Properties of convective features are also determined by the algorithm including finding a centroid and attempting to fit an ellipse to define the structure. Additional properties of convective features are determined within the classification scheme.