THE IMPACT OF SPATIAL RESOLUTION ON RECONSTRUCTION OF SIMPLE PATTERN THROUGH MULTI LAYER PERCEPTRON ARTIFICIAL NEURAL NETWORK

سال انتشار: 1396
نوع سند: مقاله کنفرانسی
زبان: انگلیسی
مشاهده: 550

فایل این مقاله در 7 صفحه با فرمت PDF و WORD قابل دریافت می باشد

این مقاله در بخشهای موضوعی زیر دسته بندی شده است:

استخراج به نرم افزارهای پژوهشی:

لینک ثابت به این مقاله:

شناسه ملی سند علمی:

CITCOMP02_230

تاریخ نمایه سازی: 7 اسفند 1396

چکیده مقاله:

This study investigates the effect of spatial resolution factor on prediction of symbol ‘I ‘, the first Roman number, from scatter pixel images of this symbol through multiple perceptron neural network with adjusted network training parameter. The corrupted images of this symbol with different spatial resolution and pixel size has applied to the feed forward neural network model. The results of modelling have revealed acceptable correlation coefficient values with low and high resolution desired images and low and high resolution predicted images. From other aspect, a very less difference between these correlation coefficients emphasize on insignificant consequence of pixel size variation over the symbol with simple geometry

کلیدواژه ها:

Artificial Feed Forward Neural Network ، Multilayer Perceptron ، Pattern Recognition ، Spatial Resolution

نویسندگان

Pardis Jafari

Department of Computer Engineering, Persian Gulf University, Boushehr, Iran

Saeideh Sarmadi

Department of Computer Engineering, Persian Gulf University, Boushehr, Iran