DeepAI AI Chat
Log In Sign Up

Multi-Temporal High Resolution Aerial Image Registration Using Semantic Features

by   Ananya Gupta, et al.
The University of Manchester

A new type of segmentation-based semantic feature (SegSF) for multi-temporal aerial image registration is proposed in this paper. These features encode information about temporally invariant objects such as roads which help deal with the issues such as changing foliage that classical handcrafted features are unable to address. These features are extracted from a semantic segmentation network and show good accuracy in registering aerial images across years and seasons.


page 6

page 7


Predicting Ground-Level Scene Layout from Aerial Imagery

We introduce a novel strategy for learning to extract semantically meani...

Joint Facade Registration and Segmentation for Urban Localization

This paper presents an efficient approach for solving jointly facade reg...

Geolocation of an aircraft using image registration coupling modes for autonomous navigation

This paper proposes to study an alternative technology to the GPS system...

Groupwise registration of aerial images

This paper addresses the task of time separated aerial image registratio...

Temporally Distributed Networks for Fast Video Segmentation

We present TDNet, a temporally distributed network designed for fast and...

Aerial Map-Based Navigation Using Semantic Segmentation and Pattern Matching

This paper proposes a novel approach to map-based navigation system for ...