Land Classification in Satellite images using Deep Learning approaches

نوع: Type: thesis

مقطع: Segment: Masters

عنوان: Title: Land Classification in Satellite images using Deep Learning approaches

ارائه دهنده: Provider: fatemeh omidikhah

اساتید راهنما: Supervisors: Dr. Hassan Khotanlou

اساتید مشاور: Advisory Professors:

اساتید ممتحن یا داور: Examining professors or referees: Dr. mouharam mansorizadeh Dr morteza heidari mozaffar

زمان و تاریخ ارائه: Time and date of presentation: 2022

مکان ارائه: Place of presentation: online

چکیده: Abstract: In recent years, machine learning problems like image classification are in the most complex problems in artificial intelligence field. One of these problems is about satellite images classification which has ways to solve since it's observed. In this thesis, deep learning techniques choosed to solve this problem in a different way with different results. Two approache based on transfer learning was proposed. First approahe is based on VGG19-Net and convolutional auto-encoder which trained immediately. The VGG-Net extracted features from convolutional auto-encoder representation and fedding into a CNN block which can classify them into categorises. Second one is based on Efficient-Net and the first part of convolutional auto-encoder. In this, the auto-encoder trained first and after that the middle layer of auto-encoder link into the first layer of Efficient-Net.

فایل: ّFile: Download فایل