ANISOTROPIC REFLECTANCE CORRECTION describes a condition where materials imaged within a scene exhibit different radiometric values when sensed from different view directions, because their reflectance varies as a function of view and solar angle. Vegetation and bare ground cover generally have a higher reflectance in the backscatter direction, which causes a trend in brightness from one side of the scene to the other. When the sun is high in the sky (i.e., small solar zenith), which is commonly the case during remote sensing acquisitions, a "hot spot" is often present in the image where the reflectance is substantially higher than for comparable surrounding features. For quantitative analyses such as image classification and change detection, these resultant variations in at-sensor radiance can cause artifacts in the resulting products.

The empirical anisotropic reflectance correction process generally reduces across-frame anisotropic reflectance by 50-100% and largely removes the effect of hot spots from image frames. The correction procedures were utilized to successfully normalize frame imagery for change detection analysis. Examples of the pre- and post-correction images and change prodects are shown below.
The images from pass 1 and pass 2 were corrected using the anisotropic reflectance correction procedures. It is evident that the correction procedure substantially reduced the hot spots within the images. It is also apparent in the difference images that the correction process substantially reduces change detection artifacts caused by differences in hot spot locations on the multitemporal images. In addition to correcting hot-spot brightness anomalies, the model corrects within-frame anisotropic reflectance and reduces between-frame radiometric differences resulting from anisotropic reflectance.
 
Correction of anisotropic reflectance and backscatter hot spot within ADAR 5500 image fames resulting effect on difference image change detection products.