Automatic Annotation of Change Detection Images.

Nathalie Neptune,Josiane Mothe

doi:10.3390/s21041110

Abstract

Earth observation satellites have been capturing a variety of data about our planet for several decades, making many environmental applications possible such as change detection. Recently, deep learning methods have been proposed for urban change detection. However, there has been limited work done on the application of such methods to the annotation of unlabeled images in the case of change detection in forests. This annotation task consists of predicting semantic labels for a given image of a forested area where change has been detected. Currently proposed methods typically do not provide other semantic information beyond the change that is detected. To address these limitations we first demonstrate that deep learning methods can be effectively used to detect changes in a forested area with a pair of pre and post-change satellite images. We show that by using visual semantic embeddings we can automatically annotate the change images with labels extracted from scientific documents related to the study area. We investigated the effect of different corpora and found that best performances in the annotation prediction task are reached with a corpus that is related to the type of change of interest and is of medium size (over ten thousand documents).

Highlights

Earth observation satellites have been capturing a variety of data about our planet for several decades, making many environmental applications possible such as change detection
We reported the recall at k, which is commonly used for text-image/image-text retrieval tasks [26], to evaluate the visual semantic embeddings
Recall, F1 score and mean Intersection over Union obtained from training the U-net model with residual network encoders (ResNet) [27] and very deep convolutional networks originally from the Oxford Visual Geometry Group (VGG) [29], which are among the state-of-the-art convolutional neural networks for feature extraction from images

Summary

Introduction

Earth observation satellites have been capturing a variety of data about our planet for several decades, making many environmental applications possible such as change detection. Proposed methods typically do not provide other semantic information beyond the change that is detected To address these limitations we first demonstrate that deep learning methods can be effectively used to detect changes in a forested area with a pair of pre and post-change satellite images. The various changes that are happening on the Earth’s surface can be automatically detected by analyzing images of a given area taken at different times [6] Such techniques have been used to monitor loss and disturbances in forests [2,7], to track change in urban areas [8], and to map out areas affected by natural disasters [4,5]. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations

Objectives

Methods

Results

Discussion

Conclusion