سال انتشار: ۱۳۹۰

محل انتشار: سومین کنفرانس مهندسی برق و الکترونیک ایران

تعداد صفحات: ۷

نویسنده(ها):

Navid Razmjooy – Young Researchers Club Majlesi branch, Islamic Azad UniversityIsfahan, Iran
Peyman Moallem – Department of Electrical Engineering Faculty of Engineering, University of IsfahanIsfahan, Iran
B.Somayeh Mousavi – Young Researchers Club Zahedan branch, Islamic Azad UniversityZahedan, Iran
Mohsen Khalilpour – Department of Electrical Engineering Faculty of Engineering, University of MohagheghArdabil, Iran

چکیده:

This paper presents a new method for thresholding using an evolutionary algorithm for optimizing image threshold especially for object segmentation. This method is resistant versus harsh histograms because of suitable fitness function selecting that consists of the histogram details. Afterwards imperialist competitive algorithm is recomanded as an evolutionary method for minimizing the error function. Finally, a powerful algorithm for image thresholding is founded. The comparisons and experimental results show that this system is more efficiently toward other methods; particularly Otsu’s and even former evolutionary algorithms like GA.