An Invariant Recognition Method of Digital Figures with Multiple Gray Levels
Tetsuo Sagara, Teruyuki Kaneko, Takashi Takeda · Medical Entomology and Zoology · 2001
The authors have develoved a method, called a central distance method, for recognition of digital binary figures by using a “central distance function” in an earlier paper[1], which allows for shift, scaling and rotational variation. The central distance function is a kind of distribution function of distances between points consisting of a digital figure (binary valued figure) and their gravity point. In this paper we extend the central distance method to apply to digital figures with multiple gray levels (digital gray figures, for simplicity). A digital gray figure is recognized to be a set of finite number of three dimentional points by adding a third coordinate representing the value of gray levels (Fig.1). In the three dimentional space, we can introduce the central distance concept as well, and can define the central distance function as is the case for binary digital figures.