Computer Vision System for Analysis of Geometrical Characteristics of Agricultural Products and Microscopic Particles(II) -Algorithms for Geometrical Feature Analysis-

농산물 및 미립자의 기하학적 특성 분석을 위한 컴퓨터 시각 시스템(II) -기하학적 특성 분석 알고리즘-

The aim of this study is to develop a general purpose algorithm for analyzing geometrical features of agricultural products and microscopic particles regardless of their numbers, shapes and positions with a computer vision system. Primarily, boundary informations of an image were obtained by Scan Line Coding and Scan & Chain Coding methods and then with these informations, geometrical features such as area, perimeter, lengths, widths, centroid, major and minor axes, equivalent circle diameter, number of individual objects, etc, were analyzed. The algorithms developed in this study was evaluated with test images consisting of a number of randomly generated ellipsoids or a few synthesized diagrams having different features. The result was successful in terms of accuracy.