A novel approach is proposed for speckle reduction in multilook full polarimetric SAR images. In contrast to others, this approach adopts an enhanced structure detection method to estimate the parameters of the polarimetric covariance matrix for the multilook polarimetric whitening filtering (MPWF) algorithm and thus a structural adaptive and optimal speckle filter is developed. To evaluate the present approach, NASA SIR-C/XSAR, L band, four-look processed polarimetric SAR data of the Tian-Mountain Forest is used for simulation. Experimental results demonstrate the effectiveness of this novel filtering algorithm in case of both speckle reduction and preservation of texture information. Comparisons with other methods are also made.