Abstract: A model for a previously defined form is provided in terms of directional features for particular evaluation points and shift vectors from one evaluation point to the next. An input image is represented by density gradients for different directional planes. Shift operations for a counter plane which is used to accumulate values for evaluating the similarity between the image and the model, and add operations in which values from input directional planes as specified by the model are added to the counter plane, are performed repeatedly by parallel processing to find the total match with the model for the form.

Abstract: A pattern reading system by line segment approximation comprising the steps of tracing the contour and simultaneously, seeking out as candidate extreme points the points at which the inner products of coordinate point vectors and directional vectors at coordinate points of the contour being traced are largest, and feeding out these candidate extreme points as real extreme points when the differences between the inner products of the direction vectors and the inner products of the candidate extreme points are greater than an allowance set in advance.

Abstract: A method for character reading requiring no character segmentation, which method effects required character recognition by the steps of subjecting a given string of character patterns to continuous scanning to produce either a local feature vector at each of the intersections of rows and columns of character patterns or a global feature vector for each of the columns formed in consequence of the scanning, linearly consolidating either or both of the feature vectors to obtain a lower dimensional vector in new feature axes and continuously matching the vectors with the standard ones set in advance.