Mühendislik Fakültesi / Faculty of Engineering
Permanent URI for this collectionhttps://hdl.handle.net/11727/1401
Browse
2 results
Search Results
Item A Learning-Based Resegmentation Method for Extraction of Buildings in Satellite Images(2014) Dikmen, Mehmet; Halici, Ugur; https://orcid.org/0000-0002-0584-5577; AAG-8859-2019This letter introduces a new method for building extraction in satellite images. The algorithm first identifies the shadow segments on an oversegmented image, and then neighboring shadow segments, which are assumed to be cast by a single building, are merged. Next, candidate regions where buildings most likely occur are detected by using these shadow regions. Along with this information, closeness to shadows in illumination direction and spectral properties of segments are used to classify them as belonging to a "building" or not. Then, a resegmentation is performed by merging only the neighboring segments, which are classified as building. Finally, postprocessing is performed to eliminate some false building segments. The approach was tested on several Google Earth images, and the results are found to be promising.Item Early and Late Level Fusion of Deep Convolutional Neural Networks for Visual Concept Recognition(2016) Ergun, Hilal; Akyuz, Yusuf Caglar; Sert, Mustafa; Liu, Jianquan; 0000-0002-7056-4245; 0000-0002-7056-4245; B-1296-2011; D-3080-2015; AAB-8673-2019Visual concept recognition is an active research field in the last decade. Related to this attention, deep learning architectures are showing great promise in various computer vision domains including image classification, object detection, event detection and action recognition in videos. In this study, we investigate various aspects of convolutional neural networks for visual concept recognition. We analyze recent studies and different network architectures both in terms of running time and accuracy. In our proposed visual concept recognition system, we first discuss various important properties of popular convolutional network architecture under consideration. Then we describe our method for feature extraction at different levels of abstraction. We present extensive empirical information along with best practices for big data practitioners. Using these best practices we propose efficient fusion mechanisms both for single and multiple network models. We present state-of-the-art results on benchmark datasets while keeping computational costs at low level. Our results show that these state-of-the-art results can be reached without using extensive data augmentation techniques.