Abstract
This article reviews some of the problems which have been encountered in the application of automatic image analysis to problems in biology. Some of the questions involved in the actual formulation of such a problem for this approach are considered as well as the difficulties in the analysis due to lack of specific contrast in the image and to its complexity. Various practical methods which have been successful in overcoming these problems are outlined, and the question of the desirability of an opto-manual or semiautomatic system as opposed to a fully automatic version is considered.