Abstract: In order to provide an accurate sort into many different classes, objects are dropped through a viewing zone and viewed in the same instant by a number of viewers. Each viewer provides electronic signals representative of three basic shape features, namely blockiness, symmetry and convex hull deviance, together with a total count of edge-breakthroughs. For each basic shape feature, the maximum, minimum and mean are obtained for all the viewers, except the minimum convex hull deviance signal. These signals, together with the edge-breakthrough count, are each subjected to a linear transformation to provide a normalized shape parameter which is then assigned a value of 0 to 15 for each class being sorted on the basis of the expected occurrence of that parameter in that class. Each pair of secondary shape parameters so determined is used to derive from a respective table a decision value. The shape class of the object is ascertained on the basis of a majority vote for all the shape decision values.