Title :
One extraction method of morph character parameters based on contour tracking
Author :
Ding Youchun ; Chen Hong
Author_Institution :
Sch. of Eng. & Technol., HuaZhong Agric. Univ., Wuhan, China
Abstract :
Aiming at dealing with the complexity of connected component labeling conflict, one labeling method based on contour tracking is put forward. Using different markers for different objects and inner and outer contour solves multi-objects contour tracking effectively. Through scanning the binary image once, morph character parameters such as area, perimeter, form-center, long diameter and short diameter, roundness and rectangleness, are extracted accurately. The application in the duck eggs´ quality detection indicates that this method is good for extensive use.
Keywords :
computer vision; binary image; contour tracking; extraction method; labeling method; morph character parameters; multiobjects contour tracking; Bismuth; Character recognition; Computers; Image edge detection; Labeling; Pixel; Target tracking;
Conference_Titel :
World Automation Congress (WAC), 2010
Conference_Location :
Kobe
Print_ISBN :
978-1-4244-9673-0
Electronic_ISBN :
2154-4824