Least square method for enhancement of laser radar images based on piecewise linear transformations of gray scales

Images of laser radars for meteorological measurements need enhancement by gray scale transformations to have outputs of good quality. Ordinary techniques of the enhancement such as the histogram flattening use nonlinear transformations. The laser radar images, however, prefer linear transformations, since the nonlinearities may induce unexpected deformations on the enhanced image and the outputs become unreliable. This paper is concerned with the problem of matching between the histogram of a transformed image and a prescribed histogram by optimizing a criterion of fitness of two histograms. A least square method is considered as one of such criteria, and is minimized by controlling parameters which describe a class of piecewise linear gray-scale transformations. Although the present method is developed for the laser radar images, it is applicable to various kinds of pictures obtained by scientific measurements.