Georeferencing images is a laborious process so schemes for automating this process have been under investigation for some time. Among the most promising automatic registration algorithms are those based on the Fast Fourier Transform (FFT). The displacement between two given images can be determined by computing the ratio F 1 conj(F 2 )/|F 1 F 2 |, and then applying the inverse Fourier transform. The result is an impulse-like function, which is approximately zero everywhere except at the displacement that is necessary to optimally register the images. Converting from rectangular coordinates to log-polar coordinates, shifts representing rotation and scaling can also determined to complete the georectification process.Our FFT-based algorithm has been successfully implemented in IDL (Interactive Data Language) and added as two user functions to an image processing software package -ENVI (ENvironment for Visualizing Images) interface. ENVI handles all pre-processing and postprocessing work such as input, output, display, filter, analysis, and file management. To test this implementation, several dozen tests were conducted on both simulated and "real world" images. The results of these tests show advantages and limitations of this algorithm. In particular, our tests show that the accuracy of the resulting registration is quite good compared to current manual methods.

No abstract

-This paper describes a technique for automating the measurement of brain width and length at the level of the hi-parietal diameter, by processing an axial computed tomography (CT) brain scan image. The development of this algorithm derives from the wish to normalise patient data according to skull size and.shape, for the purpose of comparing new patient data with that fiom past cases. This algorithm uses image processing techniques to find the inner edge of the cavity of the skull. The width and length of the brain are measured as inner dimensions of this hone periphery. The main challenges facing this work are the structural asymmetry of the brain and the angle of rotation commonly encountered whcn working with axial CT images. Both of these must be taken into account prior to measuring the brain width and IengIh. The algorithm was designed and tested to operate on a database containing CT brain scans &om 530 patients. The results indicate that the algorithm bas a 90S6% success rate.